Dresden Nuclear Power Station, Units 2 and 3, Response to … · 2017-07-31 · Please note that Irfan Hussain, whose resume is included in Appendix A to the Relay Evaluation Report,

RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION RE UNRESOLVED SAFETY ISSUE (USI) A-46

REC'D W/LTR DTD 04/11/98 .... 9804160292

-NOTICE-THE ATTACHED FILES ARE OFFICAL RECORDS OF THE OCIO/INFORMATtON · .

~, ' iv1Ai~A\.:it:1V1t:.NT DIVISION. THEY HAVE BEEN CHARGED TO YOU FOR A LIMITED TIME PERIOD AND MUST BE RETURNED TO THE RECORDS AND ARCHIVES SERVICES SECTION, T-5C3. PLEASE DO NOT SEND DOCUMENTS CHARGED OUT THROUGH THE MAIL. REMOVAL OF ANY PAGE(S) FROM DOCUMENTS FOR REPRODUCTION MUST BE REFERRED TO FILE PERSONNEL.

-NOTICE-

\

\ \ '· \ \

,•

NRC Question #1

ATTACHMENT A RESPONSE TO QUESTIONS

In Appendix I to the Seismic Evaluation Report attached to the referenced letter, the licensee listed all of the USI A-46 outliers for equipment, heat exchanger, and cable trays. In Appendix G to the Relay Evaluation Report, the licensee also listed all of the _relay outliers. In Note 1 on each page of both Appendixes, the licensee indicated that all outlier resolutions, either by analysis, physical modifications, or replacements will be completed for each respective unit by the end of the second refueling outage for that unit after receipt of the NRC stafPs Safety Evaluation. The licensee is requested to elaborate on its d~cision to defer the resolution of identified outliers and its evaluation in support of the conclusion that the licensing basis for the plant will not be affected by its decision. Specifically, the licensee is requested to provide the justification for assuring operability of the affected systems and components while a number of safety-related components in the safe shutdown path have been identified as outliers, thus, rendering their seismic adequacy · questionable and their conformanGe to the licensing basis uncertain.

ComEd Response t.o Question #1

The response to this question is provided in three parts.

In response to "The licensee is requested to elaborate on its decision to defer the resolution of ,_ identified outliers ... ", we prov\de the following:

· In the summary report submittal, Dresden commit~ed to "resolve all of the outliers for each re~pective l]nit by the end of the second refueling outage for that Unit after receipt of the Safety Evaluation Report [SER]." This commitment was made based on SQUG guidance that two

. I .

refueling outages was an acceptable time span and in anticipation that the SER w~s forthcoming.

It was not Dresden's· intention that the summary report submittal convey we were deferring outlier resolution. Following the submittal, Dresd.en began activities toward resolution of the outliers. Approximately 40 % of the equipment outliers have been resolved (closed). Dresden is currently pursuing resolution of the remaining outliers.

With respect to relay outliers, the plan is to initiate additional analysis, circuit evaluations and industry data research with resulting modifications, as required, thereafter, consistent with ComEd's original commitment for outlier reso!Ution.

In response to "; .. and its evaluation in support of the conclusion that the licensing basis for the plant will not be affected by its decision." ComEd provides the following:

With regard to assuring operability of systems and compc;ments, D~esden has followed guidance provided in the Generic Implementation Procedure (GIP). Section 2.2.5 of the GIP describes in detail the guidelines used by Dresden in implementing its responsibility to reporting under the USI A-46 program.

GIP Part I, Section 2.2.5 states:


"failure of equipment to meet GIP initial screening or outlier resolution guidelines does not, of itself, give rise to a need for the licensee to consider a JCO or reporting under applicable reporting requirements unless the plant has modified its commitments to adopt the USI A-46 (GIP) methodology as its licensing basis for verifying the seismic adequacy of electrical and mechanical equipment, as set forth in Paragraph 2.3 .3 below. If a determination is made' that equipment failure to meet the GIP initial screening or outlier resolution guidelines does not meet the existing plant licensing or design bases, including specific plant commitments and requirements, the licensee must consider reportability and operability implications pursuant to Technical Specifications and lOCFR 50. 72, 50. 73 and 50.9, among others as appropriate, including the need for JCO. Non-safety grade equipment selected for use in ·usI A-46 is not exempt from reporting requirements.

There is no independent requirement ·to notify the Staff ~hen an outlier or equi,pmentdeficient against USI A-46 criteria, which is not a deficiency against the plant's licensing· or design basis, wiH not be modifi~d to conform to the l;SI A-46 guidelines. However, .licensees will report unresolved USf A-46 outliers or equipment deficient against USI A-46 criteria to the NRC Staff in their final summary reports. These reports will include an explanation of the safety implications of not modifying the .outlie~s or equipment deficient against USI A-46 criteria. The Staff will justify any requirement to modify these outliers under 1~. CFR. 50.109 as. stated in Section 2.3.1 b~low."

the G~ methodology screens out equipment which passes a set of conservative, generic,'seismic adequacy criteria. Equipment which does not pass this screen is identified as .an outlier and either is resolved by one ofthe resolution methods de~cribed in the GIP (Part II, Section 5.3) or is left as an unresolved outlier. As discussed above, reportability and. operability determinations under. 10 CFR. 50.72 and 50.73 are not part of the GIP methodology and are not considered unless the unresolved outlier also represents a condition outside the plant's design basis.

In response to " ... the licensee is requested to provide the justification for assuring operability of the affected systems and components ... ", we provide the following:

' )

Dresden has reviewed all outliers against its seismic design basis in accordance with GIP methodologies .. Problem Identification Forms (PIFs) were generated for outliers which were identified by the Seismic Review Team as potential operability issues. The operabiJity of · identified components and systems was reviewed _utilizing the appropriate station operability determination procedure. . ·

2

,,<

•

••

NRC Question #2


On Page 4-3, the report stated that a total of 5638 contacts were identified during the detailed circuit analysis. Of those contacts identified, 1210 were screened as "Chatter Acceptable," 193 9 were identified as "Not vulnerable" and 52 contacts were screened from further consideration using "Operator Action." On page 4-4, the report· stated that following the detailed circuit analysis, a total of 2340 contacts was evaluated for seismic adequacy using the Seismic Capacity Screening Methodology. Provide information to show how the remaining 97 contacts were verified for seismic adequacy. · ·

Furthermore on Page 4-5, the report stated that of the 2340 contacts evaluated using the seismic · capacity screening process, 1252 of the contacts passed. The 1058 contacts, which did not pass,

have been declared outliers. Provide information to show how the remaining 30 contacts were verified for seismic adequacy.

ComEd Response to Question #2

After an additional review of the electronic database fil~, which is the basis for the Dresden relay list, we have ascertained that an· additional 85 contacts were originally reviewed. The total

I ,

number of contacts reviewed is 5723, not 5638 as reported in the"Dresden SQUG summary report submittal. Consequently the number of contacts screened for '·'Chatter Acceptable (CA)", ''Not Vulnerable (NV)", and "Operator Action (QA)" has changed. The correct contact breakdown is:

Number of contacts screened as "CA" Number of contacts screened as "NV'' Number of contacts screened as "OA" =

1259 2084 40 '

The number of contacts evaluated using the Seismic Capacity Screening process as shown on Page 4-5 is correct and is 2340. Also the number of contacts which passed this screening is correct and is 125.2. However, the number of contacts which did not pass the screening and have been declared outliers should have been 1088 not 1058 as reported in the summary report, which accounts for the additional 30 contacts. These 30 contacts were in fact reviewed originally.

· Therefore, there are no unreviewed ·or un.screened contacts.

\'

NRC Question #3

In Appendix C to the Seismic Evaluation Report, "Dresden Nucl~ar Station Walkdown Personnel Resumes," and Appendix F to the Relay Evaluation report, "Resumes oflndividuals Performing Relay Review," the staff noted that certificates were not provided for some of the personnel who participated in the seismic walkdown inspections and relay review to demonstra.te that they have completed all the necessary Seismic Qualification Utilities Group (SQUG) training courses. The licensee is requested. to provide appropriate documentation to demonstrate,that those individuals· -· are qualified to participate in the USI A-46 Implementation program.

3

·'··

• ATTACHMENT A

RESPONSE TO QUESTIONS


Attachment B provides the missing certifications for Tim Loch, George Thomas, John Stevenson, Robert Kennedy, J. Bednarczyk, and Michael Warpehoski. Please note that Irfan Hussain, whose resume is included in Appendix A to the Relay Evaluation Report, was assisting the Lead Relay Reviewer and was not required to complete the SQUG-developed relay training course. Per Sections 2.1.3 and 4.5 of GIP-2, only the Lead Relay Reviewer is required to successfully complete the SQUG-developed training course on the relay screening and evaluation procedure.

NRC Question #4

In Appendix D to the Seismic Evaluation Report, it is.not clear how the less than 40 ft criterion has been established. Provide the effective grade-elevations for each of the seismic Class 1 structures. It appears that in some cases, the seismic demand for equipment located within 40-feet above the effective grade has been defined by the Hou~ner ground response spectrum (GRS) and compared with the bounding spectrum, utilizing provision A. l of Table 4-1 of GIP-2. Section 4.2.3, "Advantage and Limitations," of GIP-2 indicates that this provision is based on an assumption that the factor between the GRS and the IRS will not be more than I . 5. A review of an IRS at 545 .ft 6 in. (N-S direction) for example, indicates that the demand for the equipment

_ located on this elevation is higher than 'l. 5 times the bounding spectrum at frequencies higher than 8 Hz. Provide justification for using this approach even when the amplified IRS is shown to be higher than the 1. 5 times the bounding spectrum and identify the floor wh~re such conditions exist. Also, explain the acronyms DOC, RRS, and CRS used in Appendix D.

ComEd.Response to Question #4

Per the Structural Design Criteria, plan.t grade is defined as 517 '-0". Dresden is a rock site. The Class 1 structures ~are embedded in or founded on rock. By the GIP .(Section 4.2.3), effective grade at a 11uclear power plant is defined as the average elevation of the ground surrounding the building along its perimeter.

Regarding how Dresden performed the ground spectrum/bounding spectrum comparison, Dresden followed the guidance as given in the GIP. Please refer to Item 1 of a recent SQUG · letter (Reference B) to the USNRC regarding the restrictions and limitations on the use of GIP Method A. ComEd has complied with the GIP and the clarifications contained in the referenced letter. -.

The acronyms DOC, CRS, and RRS are directly from the GIP, Section 4.6, and are given below:

DOC Component-Specific Seismic Qualification Documentation CRS Conservative, Design In-Structure Response Spectrum RRS Realistic, Median-Centered In-Structure Response Spectrum

4

•

••••

NRC Question #5


GIP-2 (Section 4.4) recommends that expansion anchors not be used for anchoring vibratory equipment, such as pumps and air compressors. If used, GIP-2 recommends a large margin between the pullout loads an4 the pullout capacities. The screening verification data sheets (SVDS) in Appendix D to the Seismic Evaluation Report did not provide any informatio.n regarding the type of anchors used for the listed equipment. Provide information about the seismic adequacy of vibratory equipment secured by expansion anchors.


Pumps, chillers, and air compressors on the Dresden Safe Shutdown Equipment List (SSEL) are listed below in Attachment A Table 1. The table contains 44 equipment items of which 17 items are anchored by expansion anchors. All expansion anchored equipment items, except the diesel generator cooling water pumps which have expansion anchors encased in concrete, had all of their expansion anchors checked for tightness successfully - i.e., no outliers or anomalies. As shown in the AttactmJent A Table 1, the factors of safety with respect to ultimate load capacities are large; thus, no further action is deemed necessary.

GIP Class 5

5

5

5

5

5

5

5

5

5

5

5

' '

Equipment Mark No. D02-2302-P30

D02-2301-T20

D02-2301-0057-. P30 D03-2302-P30

D03-2301-T20 ~

D03-2301-0057-P30 D03-3903-P30

D02~3903-P30

D02-5203-P30

D00-5203--P30

D03-5203-P30

· -D00-3903---P30

Attachment A Table 1 SSEL i>um11s, :Chill~rs. and Air Compressors

Description Building

HPCI/ HPCI Pump 'Reactor

'·

HPCJ/ HPCI Turbine Reactor

HPCI/ J:IPCI Turbine Cooling Water · .~eactor

Pwnp . HPCI/ HPCI Pump · Reactor

HPCI/ HPCI Turbine Reactor

HPCl/.HPCI Turbine Cooling Water Reactor Pwnp SER VICE WA TERI Diesel Generator C. House Cooling Water Pump SER VICE WA TERI Diesel Generator C. House Cooling Water Pump DIESEL GENERA TORI F~el Oil Turbine Transfer Pump DIESEL GENERA TOR/ Fuel Oil Reactor Transfer Pump DIESEL GENERA TOR/ Fuel Oil Turbine Transfer Pump -

'SERVI~E WATER/Diesel Generator··· C: House Cooling Water Pump

5

Elevation. Expansion Anchored

476' 6" No

476'6" No

476' 6'" No '·

476'6" No '

476' 6" No

476' 6~' No

490' 8" Yes'

' F.S. >30 490'8" Yes

F.S. >30 51'7' 6" No

504'6" No

517' 6" No

490"8"' Yes··· F.S. >30

GIP Equipment Mark

• Class No. 5 002,2301-

CONDPP 5 003-2301-

CONDPP 5 002-1501-

0044AP30 5 002-1501-

0044BP30 5 002-1501-

0044CP30 5 002-1501-

0044DP30 5 003-1501-

0044AP30 5 003-1501-

0044BP30 5 003-1501-

·0044CP30 5 003-1501-

0044DP30 6 003-l 502-A-P30

6 003~1502-B-P30'

6 D93-l 502-C:-P30

6 003- l 502-D-P30

6 002- l 502-A-P30

6 002-l 502-B-'-P3b

6 002- l 50'.Z-C-P30

6 D02~ l 502-D-P30

l l D00-9400-0 l 0 I-FIO

11 000-9400~ 102-Rl5

12 002-4600-B-T05

12 002-4600-C-T05

12 D02-4600-G--T05

12 D02-4600-H--T05

12 D03-4600-B----T05

12 003-4600-C--T05

12 D03-4600~H--T05

• -- --- ·- - . - - .. 0

12 D03-4600-G--T05



HPCI/ Condenser Hotwell Condensate Reactor Pump HPCI/ Condenser Hotwell Condensate Reactor Pump I

CCSW/ CCSW Pump A Tiirbine

CCSW/ CCSW Pump B Turbine

CCSW/ CCSW Pump C Turbine

CCSW/ CCSW Pump D Turbine

CCSW/ CCSW Pump A Turbine

CCSW/ CCSW Pump B Turbine

CCSW/ CCSW Pump C Turbine

CCSW/ CCSW Pump D Turbine

LPCI/ LPCI Injection Pump A Reactor

LPCI/ LPCI Injection Pump B Reactor

LPCI/ LPCI Injection Pump C Reacfor

LPCI/ LPCI Injection Pump D Reactor

LPCI/ LPCI Injec~ion 1Pump A , Reactor

LPCI/ LPCI Injection Pump B Reactor

. LPCI/ LPCI Injection Pµmp C Reactor

' LPCI/LPCI.Injection Pump D . Reactor

CONTROL ROOM VENTILATION/ Air Turbine Filtration Unit Heater CONTROL ROOM VENTILATION/ Turbine Refrigeration Condensing Unit DIESEL GENERA TOR/ ·Primary Gas Turbine Air Receiver Unit,Al DIESEL GENERA TOR/ Primaiy Gas Turbine Air Receiver Unit A2 DIESEL GENERA TORI Primary Gas Turbine Air Receiver Unit Bl DIESEL GENERA TORI Primary Gas Turbine Air Receiver Unit B2 DIESEL GENER/\ T<;>RI Primary Gas Turbine Air Receiver Unit Al · DIESEL GENERA TOR/ Primary Gas Turbine Air Receiver Unit A2 DIESEL GENERA TORI Primary Gas Turbine Air Receiver0 UnitB2 - -·

DIESEL GENERA TORI Primary Gas Turbine. Air Receiver Unit BI

6

Elevation Expansion Anchored

476'6" No

476' 6" No /

495' 0" No

495' 0" No

495' 0" No

495' 0" No

495' 0" No

. 495' 0" No

495' 0" No

495' 0" No

476' 6" No

476' 6"· No

476' 6 11 No

476' 6 11 No

476' 6" No

476' 6" No

476' 6" No

476' 6" No

534'0" Yes F.S.-9.6

534'0" Yes F.S.-19.8

528' 3" Yes F.S.-17.4

528' 3" Yes F.S.-17.4

528' 3" .Yes F.S.-17.4

528' 3" Yes F.S.-17.4

517' 6 11 Yes F.S.-17.4

517' 6" Yes F.S.-17.4

517' 6 11 Yes -- .. · TS:""l7.4

s 17' 6 11 Yes F.S.-17.4

GIP Equipment Mark Class No. 12 D00-4600-B-T05

-12 D00-4600-C-T05

12 D00-4600-G-T05

12 D00-4600-H-T05

'

NRC Question #6



DIESEL GENERA TORI Primary Gas Reactor Air Receiver Unit Al DIESEL GENERA TORI Primary Gas Reactor Air Receiver Unit A2 DIESEL GENERATOR/ Primary Gas Reactor Air Receiver Unit BI DIESEL GENERATOR/ Primary Gas Reactor Air Receiver Unit B2

Elevation Expansion Anchored

504'6" Yes F.S.-17.4

504'6" Yes F.S.-17.4

504'6" Yes F.S.-17.4

504' 6" Yes F.S.-17.4

In Table 5-1, Commentary Regarding GIP Deviations, the licensee lists some interpretations cir measures that were taken to meet the intent of the GIP caveats. For some electrical equipment identified in that Table, the Commentary states that capacity exceeds demand in a small region (small frequency.range), but in the judgement of the Seismic Capacity Engineers it was co.nsidered inconsequential. It is not clear as to how that equipment was determined to meet the i~tent of the GIP caveat. The licensee is requested to provide additional information to demonstrate that how · it was determined that the intent rather than the specific letter of the referred GIP ca\feat has been met.


For the electrical ~quipment items discussed in'.Table 5-1, Th_e Seismic Review Team (SRT) .was comprised of Seismic Capability Engineers (SCEs) Drs. Robert Kennedy and John Stevenson.

With regard to equipment items D02-7826-4----M05, D02-8302Al----P06, D02~8302B-l---P06, D02-83125-1---P06, D02-83125-2---P06, D02-83250-----M05, D02-9802-A----P06, D02-9802-B----P06 D02-9802-A----B04 D02-9802.:B----B04 D00-83250-0---BOS and 002-83250:.2---, ' ' . ' .. BOS located in the Turbine Building on Elevation 549', the demand· spectrum exceeds the capacity spectrum by less than I 0% over a frequency ·span of about 18 to 26 Hz. This was adjudged to be . .

·a small exceedence by the SR Tand iri the context of exercising engineering judgment as allowed . .

by the GIP, they concluded this equipment was acceptable.

With regard to equipment items D02-7128------T 10, 002-7229------T 10, 003-713 8-7---'-T.l 0, and 003-7239------TlO located in the Reactor Building on Elevation 570', the SCEs adjudged the fundamental frequency to be gre.ater than 11 Hz and\..also adjudged the demand/capacity exceedance above the fundamental frequency to be acceptable and in the context of exercising

· engineering judgment as allowed by the GIP, they concluded this equipment was acceptable.

NRC Question #7 ·

Table 5-1 in Section 5 indicates that the low-pressure coolant injection (LPCI) pumps are anchored by epoxy grouted anchor bolts. Recent testing sponsored by the NRC Office of

'Research at the University of Tex~s .. has shown that the dynamic behavior of the Portland cement · grouted anchors is inferior to that of the other expansion anchors and certainly inferior to that of

7

\

• ATTACHMENT A

RESPONSE TO QUESTIONS

the cast-in-place anchors (Reference C). Provide test data which could verify that the epoxygrouted anchors used for the 'LPCI pump supports have the same allowable as the cast-in-place anchors.

ComEd Reponse to NRC Question #7

ComEd is not aware of any test data for epoxy_grouted anchors, however, i.t is acceptable to use GIP capacities for cast in place anchors for the following reasons.

The subject University of Texas (UT) dissertation includes the results of tests for single concrete anchors in tension only, in cracked and uncracked concrete slabs, for static and dynamic loads. The test results show that the failure loads for grouted anchors are lower for cracked concrete than for uncracked concrete. ComEd agrees with that conclusion; however, it is not applicable to the grouted anchors used for the LPCI pumps at Dresden because: '

• there is no evidence of concrete cracking in the vicinity of the anchors for the LPCI pumps,

• epoxy grout, which normally exhibits superior adhesion properties to cementitious grout, was utilized,

• the sample anchors used in the dissertation were short (shallowly embedded), whereas the LPCI anchors are deeply embedded (11" or gre~ter).

In the dissertation, the Concrete Capacity Method (CC Method) is used to calcµlate concrete . ·' pullout strength. The equation given on pg. 14 of the dissertation is essentially the same as that given in the EPRI Anchorage Report (Reference D) on pg. 2-89 used to establish the pullout strength of cast-in-place anchors in the GIP.

The GIP minimum embedment lengths (L111J assure that the steel anchor will fail in yield . , . before the concrete fails. For embedments less than L111in, the pullout reduction formula for RLP adjusts the cast-in-place anchor capacity as a function of the embedment depth. The GIP method yields a .factor of safety of more than 4 in capacity with respect to the CC Method for . the example shown in Figure 2.2.P of the EPRI NP-5228 report for uncracked (crack width =

0.0) concrete with a strength of 3500 psi. Referring to that figure, the CC Method calculates a pullout capacity of 13.2 kips for a 1/i" diameter cast-in-place anchor with an embedment length of 3..15 in. (80 mm). The GIP requires an -embedment length of 5 in. to achieve the full strength of 6.66 kips. Applying the RLP reduction factor from the GIP for the 3.15 in. embedment length results in a calculated capacity of 3.13 kips.

Capacity of%" Diameter Cast-In-Place Anchor in Uncracked Concrete

Embedment Length GIP Allowed CC Method Capacity

3.15 in. 3.13 kips 13.2 kips .5.00irt .. . . ·· .. 6,66- kips · . -· - . 26:5 kips···· ·

8

•

•••


This data clearly reinforces the reasonableness and inherent conservatism in the GIP capacity values. It is important to point out that the UT dissertation concludes that the same concrete capacity expression as given on pg. 14 of the dissertation and value for the constant, k, ( 40 for headed anchors in US units) may be used for both cast-in-place and grouted-in-place anchors in uncracked concrete. This conclusion is given on pg. 240 of the dissertation. It is also noteworthy that the grout used in the installation of grouted anchors was a commercial, non-expansive, premixed, sanded, cementitious grout, combined with the amount of water recommended by the manufacturer. Non-shrink grouts are expansive grouts in that they use a number of different types of materials and/or chemicals to compensate for the shrinkage of the grout, which i~ due to water evaporation, to ensure that there is not a significant net volume reduction. Dresden utilized a non-shrink bonding compound, "Colma" Epoxy Polysulfide, as manufactured by Sika Cherriical Corp. (Reference E) for the LPCI pumps.

NRC Question #8

In Table 8.2, USI A-46 Equipment Outliers, it appears that in some cases, the lic~nsee has relocated spare circuit breakers due to the concerns of potential seismic spatial interactions with certain switchgears. It is not clear as to where those spare circuit breakers were located. However, it should be noted that removal of the circuit breaker from the switchgear will result in mass redistribution of the switchgear. Mass redistribution of the switchgear may then change the frequency of the switchgear ~nd its dynamic response during a seismic event and may invalidate the original seismic qualification of the switchgear. The licensee,is requested to evaluate this concern for applicability to those identified switchgear.


Spare circuit breakers for the safety-related 4KV switchgears have been relocated from the proximity of the switchgear to an approved breaker storage area on the same elevation of the Reactor Building (Ref. Qresden Operating Procedure (DOP) 6500-04). The SQUG walkdown SRT made note of any seismic interaction concerns and outliers declared as necessary due to the potential interaction between a free-standing circuit breaker and any nearby SSEL equipm~nt.

The seismic qualification design basis of the switchgear, along with most other equipment at· Dresderi station, is not well defined. Hence Dresden is a SQUG plant and underwent the rigorous review specified in the SQUG GIP as agreed to ·with the NRC.

With respect to circuit breaker removal from switchgear, there is n<;>. design basis impact on the original switchgear seismic qualification, bas~d on the above argument.

NRC Question #9

In Section 6 of the Seismic Evaluation Report attached to the referenced letter, the licensee indicated that the Con~e_ryati\;'e.Deterministic Failure Margin (CDFM) method-of Electric-Power Research~i~stitute (EPRI) NP-6041, "Seismic Margin Report," was utilized to resolve the vertical

9

j

, . . .

•


tanks that did not meet the GIP caveats. The methodology has not been approved by the staff for the analysis of safety-related systems and components, including the resolution of mechanical, electrical, and structural component outliers in the USI A-46 program. The licensee is requested · to reevaluate its program and to ensure that all the identified systems and components will be resolved using the plant licensing-basis methodologies or other approaches acceptable to the staff

ComEd Reporise to Question #9

The Conservative Deterministjc Failure Margin (CDFM) Method of Electric Power Research Institute (EPRI) NP-6041, !'Seismic Margin Report" was only used for the Contaminated Condensate Storage Tanks (D00-3303-A----TOS & D00-3303-B----TOS). It was utilized only after the Contaminated Con_densate Storage Tanks (CCSTs)' had been declared outliers by the

·GIP process. As discussed in the answer to Question 10, the CCSTs were reanalyzed in accordance.with th_e GIP methodology and shown to be acceptable. The CDFM or "Margins" approach has not been used for any other equipment at Dresden.

NRC Question #10

Section 6 of the Seismic Evaluation Report provides a summary describing the methodology for evaluating large, flat-bottom, vertical tanks (two c,0ndensate storage tanks). However, Table 6.1 does not contain information ~bout either of them. Provide the follow,ing information about these tanks:

a. Sketches showing tank dimensions, anchor chairs, anchorages (including·embedment) and foundation.

b. A detailed calculation ofa representative tank in accordance with the GIP;..2 procedure. If it can not meet the GIP caveats, describe the weaknesses; including an assessment of potential modifications, that if implemented,. could enable the tanks to meet the GIP-2 criteria.

c. Calculations showiryg the adequacy of the ring foundations. '·


Calculation 93C2806A:C-003 pertaining to the two Contamin.ated Condensate Storage Tanks (D00-3303-A----TOS _& D00-3303-B----TOS) is enclosed.' Also enclosed are relevant drawings for the tank and its foundation, as requested. The calculation shows· that the HCLPF is 0.26g

I

peak ground 'acceleration (PGA) with respect to the SSE design basis input using 5 % damping ·and 0.9 buckling knockdown factor in accordance with the procedure in Appendix Hof the EPRI NP-6041 Seismib Margins Report. The Dresden site design basis PGA is 0.20g.

Page 10 of Stevenson and Associates Calc~l~tLon C-003 contains th~ evaluation of the. r:ing -. fooilaation in ac-fordance wifh j~:PRI-TR-103960 (Reference F) and was shown to be acceptable.

10

•


A new calculation 98Q4006-C-002 (also enclosed) recalculates the tank capacity in accord~nce with the GIP rules using the methodology and factors as given in Section 7 of the GIP. This new calculation shows a tank PGA capacity of 0.2lg, which exceeds the Dresden site design basis PGA of0.20g; thus, the Dresden Condensate Storage Tanks seismic capacity exceeds the site SSE design basi~ and the tanks are no longer outliers.

NRC Question #11

Item 5 of Table 6-1 indicates that ComEd is currently (when the Summary 1~.eport was developed) evaluating the adequacy of the fo~r LPCI heat exchangers' support steel. Provide a summary of the results of the evaluation and details of any proposed or implemented modifications on the support steel.


ComEd evaluated the C'.ldequacy bf the support steel for the four LPCI heat exchang~rs located in four LPCI corner rooms. The eyaluation consisted of a computer analysis of the support steel with all the applicable loads from the heat exchanger and the connected piping. The results indicated that the support steel in all four corner rooms did not meet all of the design

· basis requirements stipulated in the Dresden UFSAR. Necessary .modifications were implemented to bring all portions of the support platform to within the design ~a~is requirements.

The modifications consisted of adding additional structural' members and reinforcing numerous existing connections to provide additional vertical and horizontal support to the heat exchangers. These modifications were performed under Modification Num.bers E12-2-95-242 · and E12-2-95-243 for Unit 2 ·and E12-3-95-258 and E12~3-95-259 for Unit J.

NRC Question #12

The limited analytical reviews'(LAR) described in Section 7.3 indicated that 10 out of 12 reviews· required outlier evaluations. In light of this finding, provide a justification for not expand.ing the -reviews to a larger sample size, particularly, when the outlier evaluations indicated hardware modifications.


The Limited Analytical Reviews (LAR) in Section 7.3 of the USI A-46 final report for Dresden determined that 10 out of 12 reviews were acceptable .. Five were acceptable outright meeting all of the GIP capacity checks (see Table 7 .2 in the Dresden USI A-46 Summary Report). Another 5 LARs were declared outliers and required further evaluation analysis beyond the GIP capacity checks, but were subsequently resolved analytically without hardware modifications and are no _long~r o"utli~rs (S.eecTable .7.3 incthe-Dresden, USI A-46 Summary Rep-orif Ofthe remruning t-wo outliers, LAR 009 was awaiting confirmation of the actual

II

•

••


weight per ft. of the enclosed conduit duct raceway. The capacity of the raceway is 98 lbs. /ft. as given in Table 7.2 in the Dresden USI A-46 Summary Report. It has since been confirmed that the actual weight of the conduit duct raceway is less than 20 lbs./ft.; thus, LAR 009 is resolved and no longer an outlier. This currently leaves one remaining, unresolved LAR conce~ing cable tray raceway systems at Dresden - specifically, LAR 007. The recommendation for LAR 007 in the Dresden USI A-46 Summary Report is to expand the walkdown sample to identify additional tray configurations of that specific type for further evaluation (see Summary of Outliers in the Dresden USI A-46 Summary Report). Based on the progress thus far, Com~ is not expecting that hardware modifications will be needed to resolve this last outlier.

NRC Question #13

Provide detailed calculations showing how the outliers for LAR 001 and LAR 007 identified in Table 7.3 were resolved including a summary of the rod fatigue test data and the generic acceptability curve. ·

ComEd Reponse to Question #13

Calculations for LARs 001 and 007 are encl~sed. LAR 001 has been resolved analytically. while LAR 007 is the lone remaining unresolved outlier as previoµsly stated in the response to Question 12. The question asks for a summary of the rod fatigue data and the generic acceptability curve.· The rod fatigue data is based on the actual testing data for Dresden threaded-end rods performed for the Systematic Evaluation Program (SEP) Raceway Progr11m. Referring to Figure 8-4 in EPRI Report NP-7152-D, this data is represented as the lower . bound curves for field-threaded rods. The lower bound curves of Figure 8-4 are the. basis for the Dresden outlier evaluations. The acceptance criteria for rod fatigue was set to 50 full stress cycles for the rods whereas strong motion shaking of raceways due to actual earthquakes would not be expected to produce demands in excess of ten full stress cycles. As further justification for the use of actual Dresden ro9 fatigue. test data, we refer to the GIP rod fatigue

·· procedure that is based on the SEP rod fatigue data. Referring to Figures 8-4 and 8-5 in EPRI. Report NP-7152-D; a comparison of the SEP lower bound curves which are developed for Dresden's field-threaded rods (Figure 8-4) and the 2/3 maximum strain limited curve (Figure 8-5) on which the SQUG rod fatigue spectra are based shows they yield essentially the. same values for number of cycles to failure: . · "· · · ·

/

NRC Question #14

In evaluating the raceway supports, a recent audit of a GIP-2 plant indicated that some licensees may be misusing the "ductile support" definition of Figure 8-7 of GIP-2 to avoid the check for · later.al seismic loads. In this context, please provide the following information: (1) the number of supports (percentage of the total number of supports evaluated in the limited analytical review) considered as ductile: (2) the specific criteria used in determining that the supports are ductile; and (3) t~e exte11t to which th~ supports and their.anchorages (which do not meet the vertical · capacity check) will deform under the two horizontal components of the design basis earthquake.

12

... ;,,

•



Based on GIP-2, Section 8.3.3, Ductility Check, the total number of suspended raceway supports selected for limited analytical review is seven and all are classified as ductile. As provided in GIP-2, Section 8.3.4, Lateral Load Check, an explicit lateral load check is not required for any of these supports. At Dresden, virtually all raceway hangers are supported by threaded rod hangers. Of the total of twelve LARs, seven LAR cases are supported by threaded rod hangers. (The remaining five LARs addressed four cases of wall-mounted hanger systems and a floor-to-ceiling case ) All of the threaded rod hanger supports are ductile. This has been demonstrated by the SEP Raceway Program which used actual Dresden threaded rods in its testing programs as discussed in the response to Question 13.

In the first part of this qµestion, the RAI question asks for a percentage of raceway supports which are classified as ductile compared with the entire population of raceway supports in the plant. We did not. collect this information during the plant walkdown since ii is not part of the GIP guidelin.es. As provided in GlP-2, Section 8.2.1, the purpose of the plant walk;down is to:

"l) verify that the cable and c<:mduit racew(!y systems meet the Inclusion .Rules given in Section 8.2.2; · '

2) note and evaluate any Other Seismic Performance Concerns gi~en in S~ction 8.2.3;

3) select a sample of representative worst-case raceway supports as described in

4)

· Section 8.2.4; and

judge ~hether there are any seismic spatial interactions which could adversely affect the performance of the raceway system as outlined in Section 8.2.5." (GIP-2, pg. 8-8).

As provided in GIP-2, Sectio11 8.2.4, it is necessary to select a limited number (e.g., 10 to 20) of · the worst-case sa~ples from the plant for the Limited .Analytical Review. All vital areas were walked down by the Seismic Review Team (SRT). The LARS chosen are representative and bounding- i.e., worst case examples. Expansion of the sample size would not lead to discovery of other different outlier types or provide additional lessons learned since these examples comprise bounding cases of the cable tray hanger types at Dresden station.

In conclusion, LARs were selected for worst case examples for each type of cable tray hanger. Resolutions will be applied to hangers that fit that outlier hanger type, as needed. Therefore, there is no need to expand the reviews to a larger sample size.

In the second part to this question, the RAl asks for the specific criteria used in determining that the supports are ductile. The methodology for ductility determination is given in.GIP-'.?, Sec_tion 8.3 .3.· This is not an· issue for Dresden which predominantly has rod ·hangers support systems.

13

•

•


Rods are inherently ductile as demonstrated using actual Dresden rods in the SEP Raceway Program in which the Dresden rods were tested to establish fatigue life (see response to Question 13).

In the third part to this question, the RAI asks " ... the extent to which the support~ and their anchorages (which do not meet the vertical capacity check) will deform under the two horizontal components of the design basis earthquake." The ductility of supports for raceways in the longitudinal direction is addressed using the procedure in GIP-2, Section 8.2.3, where racew~y systems are evaluated for "hard spot" supports. No other analyses for longitudinal forces are required for raceway supports hung from above or attached to a wall. This is not an issue for Dresden since the rods themselves are axisymmetric and inherently ductile in any direction.

NRC Question #15

Tables 8.1 and 8.2 ·in the Seismic Evaluation Report show a number of equipment that do not meet the seismic demand. The special review team (SR T) has recommended methods for resolving these outliers. Provide a table showing how these outliers were actually resolved ..

ComEd Response to Question 15 \

c.omEd has provided the current outlier resolutio'n status within Attachment H.

NRC Question #16. /

Item Number A-46 in Table 8.3 in4icated ~hat the eight component cooling service water (CCSW) Pump Cooler Supports are ·anchored using lead cinch anchors. It is also noted that the licensee utilized RTR 2661 (Westinghouse Report documenting the lead cinch anchor tests for

- Savannah river site) in arriving at the capacity of the anchors: Provide information regarding the use of this test report ~ncluding the allowables and proof torque requirement, etc.

Com Ed Response to Quest_ion #16

The Westinghouse Report RTR 2661 documenting the lead shell expansion anchor tests and -criteria for Savannah River Site was used. This data, including the extreme value analysis performed by Westinghouse, was reviewed by the SCEs and it was concluded that the recommended working loads in the report were suitable for use at Dresden. At Dresderi the CCSW pump vault room coolers, which total eight in number, are anchored by lead shell expansion anchors. These coolers are suspended from brac.ed steel frames anchored to the concrete ceiling and were not tightness checked. Th~s is in compliance with the GIP, Section 4.4.1, which states that,

"This check [the torque test] is merely meant to provide reasonable assurrance that the expansion anchor is ncit loose in the hole due to gross installation defects."

14

'\

ATTACHMENT A RESPONSE TQ QUESTIONS

As further stated in GIP Section 4.4.1,

"It is not necessary to perform a tightness check of expansion anchors which are loaded in tension due to deadweight, since the adequacy of the anchor set is effectively proof-tested by the dead weight loading."

The NRG has requested the values used by Dresden for Lead Shell Anchors. Dresden utilized the values designated as "allowable load" (in bold) given in Table 2 from the RTR 2661 report. The values from that table are reproc;l,uced below:.

Lead Shell Expansion Anchors

Load, Lbs. Size, inches 3/8 1/2 5/8 3i4 1 Tension Tests

No. of Tests 35 34 17 10 9 Mean Failure Load 2623 3746 4456 5870 14058 B8M Allowable Load 7QO 1200 2000 3000 5500 (See Note A) New Tensile 600 870 970 1280 3160 Allowable Load (See Note B)

Torque Load, ft-lb· 24 . 37 .50 69 (See Note D) (See Note C) __ ,

Shear Tests No. ofTests 16 12 ., 8 (See Note E) (See Note E) Mean Failure Load 2796 5736 7934 . (See Note E) (See Note E) Shear Allowable 400 800 1400. 2000 3500 Load (See Note F) '·

Table Notes: A Savannah River Site 88M Standard values which are based in part on manufacturer's recommendations

and .which were used prior to the on-site testing program. B The tensile allowable loads are chosen to maintain at least a factor of safety of 4 with respect to the

mean failure loads. They are less than the Savannah River Site 88M Standard valu~s. c Proof torque values to ensure that Savannah River Site 88M Standard values are met .

D No uncorroded 1 inch anchors could be found to ~evelop a proof torque correlation relationship .. :

E Tests were not performed on the~ and 1-inch anchors because none•were found in the "R" R~actor Building where the testing was performed.

F The shear allowable loads are based on the Savannah River Site 88M Standard values.

\

. In closing, all of the CCSW pump vault room cooler anchorages were determined to be adequate based on the use of the anchor allowables (in bold) shown in the Table above.

15

'

ATTACHMENT B CERTIFICATION OF QUALIFIED INDIVIDUALS

. .J

.--.:.:.... ·----·

S·~a~u::: -~G ---. --· -·-~· --- --- ·--- . -. · ... --:-: -·- ~ ·-·~ -·· . -..... ---·

-

Qlrrtifirutr of Arqiru rmrnt illqis ts to C!tl'rtify iqnt

qas C!lomplrtrh tqr §~Qll<!b Walkbown §rrrrning anb §ri.smir 1£oaluation C!J:raining <!Loursr

~rlb April 6-lllt 1992

.. . ...,..

David A. Freed, MPR Associates SQUG Training Coordinator

- -

Nell P. Smith. Commonwealth Edison SQUG Ch8JJrman

Jf, (~ f!(1/.J<l·r:f,t_,rj1.vt._

Robert P. Kassawara. EPRI SQUG Program Manager

.:... - . ·~:· ~ -·- ..... -~ . ··~ -

£2;J) o.M~ SOUQ Training Cbordfnalor .

A:lber1 P. Katsewara, EPAI SOUG Program Manager

<!1rrtifiratr nf At!Jirurmrnt Qt~is is to <arrtify t~at

qas C!lamplrlrb l~r §~QI~ UBalhbown g,rrrrning anb ~rismic 1£unluution. qJ:raining C!toursr

~rib Jlunr 22-26, l 992

~X42= David A. Freed. MPR Assoclalu SQUC Training Coordinator

Neil P. Smllh, Commonweellh Cdison SQUC Chairm11n

f.t!~ Robert P. Kusawara. EPRI SQUC Proarem Manager

•

SQUG

..

··•·· <ttertificnte. nf i\c4itutmeut. . m~i.e ts tn <lrerttfy t~ut .

. .. JJtrry ithuarri!yk ... · .... ~ns Qtoritplfte~· t4.e

.&<{)lll<6 Equipment 8tlectintt au~ .lelay £ualuattitn UJrutning C!tuurse .. · ·. 4

{ijtlb Nnuemhtr 16-18, l!T!Tl

·~· ~e~~~1~ · • >· ====

SQUG.

·· <l!trtiftraft nf i\r4ttutm.eut .·. · .. UJ~is is tn C!ttrttfy t~at.

:!lirijntf C!L DnrptlJnskt .·•··· •.. ··•··· ·. •. ·. · . .·. . qas <!tnmpleteb t~e . . . . . : ·. . ..

• . • . . . . . • . • . . . ; . . f :· · 8<n'IC6 £qutpment t;elrrtton anb lelny £ualuatton i!Jratntng aruur.ae . · · · il;tlb Nnuember 1_6-1 B, 1!193 · ·

Dresden Nuclear Power Station, Units 2 and 3, Response to … · 2017-07-31 · Please note that Irfan Hussain, whose resume is included in Appendix A to the Relay Evaluation Report,

Documents