ML17199U620

Evaluation Rept on Flued Head Anchor Design

ML17199U620
Person / Time
Site:	Dresden, Quad Cities, 05000000
Issue date:	02/29/1988
From:	Bandyopadhyay, Grossman W BROOKHAVEN NATIONAL LABORATORY
To:
References
NUDOCS 8803250189
Download: ML17199U620 (14)

Text

..

EVALUATION REPORT ON FLUED HEAD ANCHOR DESIGN DRESDEN UNITS 2 & 3 and QUAD CITIES UNITS 1 & 2 COMMONWEALTH EDISON Co.

Audit Location:

Sargent & Lundy, Inc. and Impel! Corp.,

Chicago, Illinois Dates:

January 26-28, 1988 BNL

Participants:

Kamal K. Bandyopadhyay Walter Grossma.n Brookhavan National Laboratory Structural Analysis Division Upton, NY 11973 February 1988 Ir+

BACKGROUND Recently concerns were raised by the NRC regarding the operability of Dresden Units 2 and 3 Quad Cities Units 1 and 2 as a result of the discovery in September 1987 that all of the flued head containment penetration anchors were not included in the licensee's evaluation required by IE Bulletin 79-14.

In a meeting held on September 14, 1987 with cognizant NRC staff, the licensee presented the original design of the anchors and justification for treatment of the containment penetrations.

The licensee concluded that the original design, considering rupture loads, was acceptable for the flued head anchor under current design loads.

The NRC staff concluded that the presentation did not provide sufficient detail for the NRC to agree or disagree without further study of the details.

The further study was to consist of an examination of the requirements in the FSAR and an audit of the design calculations.

The support structures for the flued head penetrations of the subject units had been analyzed on the assumption of pinned connections although in reality most of these connections would transfer moments such that these structures would behave more as rigidly connected frames rather than pin-connected trusses.

In the meeting held in September 1987 it was agreed that Commonwealth Edison Company (CECo) with Sargent & Lundy (S&L) would investigate the effect of redistribution of the load if the frame connections were more realistically modelled.

In addition, it was expressed that the as-built penetration support structures should be verified for conformity with the design drawings.

Also, there were some concerns about the loads and load combinations.used in the original design co~pared to those used in recent calculations being performed by Impel! Corp.

In order to better understand the above concerns, the NRC staff have requested BNL to audit the design calculations of S&L and Impel! at their respective offices in Chicago and review their approaches in resolving the issues.

On January 26-28, 1988 BNL along with NRC staff, P-T. Kuo and J.A.

Gavula have visited S&L and Impel!.

BNL's observations and comments are provided in this report.

S&L Original Design Basis Calculations - Pinned Truss

• The original computations of a few selected penetration supports, e.g.

X-13A, X-116A, X-128 have been reviewed during the audit.

The support structure was assumed pinned and analyzed by use of the computer program STRUDL.

The frame structures were typically anchored to concrete by means of "William's Rock Bolts" which required drilling existing concrete.

Expansion anchors were also used in some instances.

The support structure was designed only for the postulated pipe break loads

• BNL offers the following comments:

1.

For the "Rock Bolts" the manufacturer recommends a tensile c~pacity equal to 50% of the ultimate load apparently in order to account for installation and other uncertainties, whereas the calculations used a capacity of 90% of the yield strength or sometimes even 90% of the ultimate strength.

An investigation should be conducted to estimate a reliable design capacity of the Rock Bolts.

Note that a safety factor of four is customarily recommended for such drill-in-concrete bolts.

2.

The allowable loads for expansion anchors could not be determined since neither the drawings nor the calculations provided adequate details of the expansion anchors, e.g. embedment length, type of anchor.

The anchor bolts installed in the field should be documented and their allowable capacity should be verified.

3.

The anchor bolts were designed only for the tensile load.

The base plates with shear lugs were used to transfer the shear loads although no calcu-lations were available to demonstrate the adequacy of the embedment to transfer the shear to concrete.

Usually the shear plate and the lugs are embedded in the cover of the concrete outside the steel reinforcements.

Test results seem to indicate looseness of concrete under high tension loads.

Therefore, the shear transfer mechanism is questionable especially in the presence of tension and shou.ld be investigated.

Note that the anchor bolts are typically weak in shear.

There was no evidence presented during the audit that the calculations were checked.

An error in inputting the structure geometry in the computer was observed for Dresden Unit 3, Penetration lllB. Therefore, CECo should demonstrate the general accuracy of the calculations.

s.

The seimsic and dead loads were not included in the design. It should be demonstrated that the penetration supports are capable of withstanding all the possible load combinations including seismic and dead loads.*

6.

The pipe rupture loa*d should be assumed to act in either direction i.e.

positive and negative.

For some load components, both.directions were not considered in the analysis. It should be demonstrated that the final design load envelopes all possible load combinations* including reversal of individual load components.

7.

The base plates were designed for the compression load only.

They should be verified for the tensile load as well.

8.

The STRUDL output shows spurious z-direction (i.e. lateral) deflection at center nodes of symmetrically loaded, symmetrical structures.

This should be investigated.

9.

The acceptance stress criteria (Fb) for the diaphragm and ring plates were different.

This should be *resolved and the appropriate allowable stress should be used for the design.

10.

For the ring plates, the allowable shear stress was assumed as 0.75Fy which apparently exceeds the yield strength in shear.

The proper allowable stress should be used.

11.

The acceptance criterion for compression members specifies an allowable compressive stress equation which is identical to equation 1.5-1 in the AISC code except for the factor of safety. The AISC specifies a F.S. which varies from 1.67 to 1.92 for main members having KL/r values from 0 to 200 (see attachment 1). The acceptance criteria has a constant F.S. of 1/0.9=

1.11.

Allowable stress values may be increased by 1.6 for analyses which include SSE and/or pipe rupture loads.

The factor of safety per the AISC equation for these analyses are then reduced by 1/1.6 and vary from 1.04 to 1.20. Using a constant F.s. of 1.1 for all n/r values results in an acceptance criterion which is below AISC requirements particularly for com-pression members with high n/r values.

This inconsistency should be in-vestigated an~ resolved since buckling of a primary compression member will result in gross structural failure of the flued head anchor truss.

S&L Recent Calculations - Pinned vs. Fixed Connections In order to investigate the load redistribution and its effect on the load-carrying capability of the penetration support structure, S&L has analyzed one frame structure (Dresden Unit 2, Penetration No. Xll6A).

The following three "different models of the same frame have been analyzed:

Model A (original design basis model):

All connections and support ends are pinned.

Model B:

Bracing connections and support ends are pinned.

The primary members are rigidly connected.

Model C:

Supports on wall and bracing member connections are pinned

• Primary members are rigidly connected.

Supports on slab are elastic and represented by springs.

The results indicate a possible increase of the member and support loads.

A local increase of the member load should not pose an overall integrity problem of the frame since in the limit the additional redundancy will further redistribute the loads ultimately approaching towards the originally analyzed pinned structure. However, the same logic may not be applicable for connections to concrete. Therefore, all concrete support connections possessing the capability of moment transfer should be assessed based on the assumption of a rigid frame.

This comment is applicable for all penetrations.

Impell Calculations It was stated during the audit that lmpell's original calculations for eight anchors were for loads other than the pipe rupture loads.

Subsequently, Impell analyzed two of these anchor structures to address the pipe rupture load (Dresden Unit 3, Penetrations X-113, X-109A).

1.

BNL offers the following comments:

Impell used the pipe break loads supplied by S&L.

Impell stated that they reduced the moment based on the plastic moment capacity of the pipe section instead of the ultimate moment previously assumed by S&L.

However, a load reduction up to six-fold was observed (e.g. X-163, Load Case II) which

2.

3.

4.

s.

could not be explained on the basis of the ultimate load alone.

CECo should verify the true load and use it in the analysis.

The shear load was neglected in the anchor design (e.g. calculation 0590-238-02, p44).

The anchor should be designed for the shear load in addition to other loads.

The analytical approach to determine the pull out capacity of Hilti expansion bolts is not reliable.

The manufacturer's recommendations along with available test data should be used to determine the possible capacity reduction due to close spacing and closeness to the edge of concrete.

The Hilti bolts in one instance (Ref. calculation 0590-238-02, p58) was observed to exceed the allowable load.

This should be further investigated.

The comments made above on the S&L calculations regarding the load combinations and design criteria are also applicable for the Impel!

calculations.

As-Built Design Regarding the as-built configuration CECo/S&L stated that several accessable penetration supports have been inspected in the field.

Most of them are very similar to the design drawings.

However, some bracing members have been observed to be missing in the field.

S&L is currently in the process of resolving the discrepancy.

The results will be soon related to the NRC staff.

CONCLUSION It is recognized that the analysis techniques and the tools available during those days when the original calculations were performed were not as sophisticated as they are today.

Therefore, assumptions were made to simplify the structures and to postulate the loads.

Some of these assumptions and loads are conservative.

On the other hand, there are some other loads or load combinations that were not included in the original design.

In addition, some of the as-built penetration support structures are reported to be different from what was analyzed and designed.

Some structural members are reportedly missing.

To date, there is no indication of verification of the as-built embedment and similar other controlling parameters for anchor bolts. Experience indicates that typically embedment and spacing of anchor bolts used in the field do not adequately conform to the design drawings necessitating a follow-up reconciliation.

In addition, documentation and tracibility of the original calculations are poor and there is evidence of a lack of coordination.

Therefore, the adequacy of the as-built penetration support structures could not be conclusively estabiished from the available information.

In conclusion, it is recommended that CECo initiate a comprehensive program to demonstrate the adequacy of the flued head penetration supports.

As a minumum, this program should include the following:

L

2.

3.

Document all as-built penetration support configurations including anchor bolts.

Verify and demonstrate reliability of the original calculations. If necessary, revise pertinent parts of the calculations.

Verify conformity of the as-built configurations with the design drawings.

If required, perform new calculations to address deficiencies.

This program should use a consistent set of design load and acceptance criteria.

Any deviation from the FSAR commitments should be justified and the pertinent part of the FSAR should be revised.

In the meeting, the comments were related CECo, S&L and Impel!.

CECo has committed to provide a comprehensive program and submit to the NRC staff for review

• Configuration Walkdown Scope The anchors listed below have significant configurational variations as identified in the January 1988 configuration walkdown and could not be dispositioned using engineering judgement.

Re-analysis is required for all FSAR load cases.

See attached walkdown procedure and checklist.

X-108A D-3 X-128 D-3 X-11 SA D-2 X-lllB D-2 X-ll 6A D-2 X-ll 6B D-2 X-123 D-2 X-124 D-2 X-144 D-2 X-36 QC-1 X-36 QC-2

\\

Bellows Replacement Scope The anchors listed below were analyzed and modified for QBE and SSE load cases in conjunction with the Bellows Replacement Program.

Re-analysis is required for pipe break load case only *. (Load combination 3).

X-l 6A QC-1 X-l 6B QC-2

8188-01 2-1-88 Page 1 Dresden/Quad Cities Flued Head Anchor Assessment Program Schedule ( //~'5 f s,/ll~j"' I /~~'-Y.ft*~ n..vfo-~-r ~

flt ;/ ~

C!lfj.--..

Item Start Q-1 Q-: 2 D-2 o.Z Q:l_

0 n~fc'-~

I.

Develop Design Criteria 2-1-88 2-8-88 2-8-88 2-8-88 2-8-88 I I.

Complete Assessment of Six RPR 1-27-88 NA NA NA 3-21-88 Anchors at Dresden - 3 I I I.

Complete Assessment for FSAR 1-29-88 4-18-88 3-7-88 NA NA Commitments of the Two Quad Cities Core Spray Flued Head Anchors Which Were Modified by Nutech IV.

Complete Assessment of the 1-25-88 NA NA 4-25-88 3-21-88 Results of the Configuration Review for Dresden

v.

Complete Assessment of the

. 1-25-88 5-30-88 4-1-88 NA NA Quad Cities Anchors (including M.S. and F.W.) using CEAs v I.

Complete Configuration Wal kdown Later Later Later Later Later If Required

SCOPE OF FLUEO HEAD ANCHOR REASSESSMENT Concrete Expansion Anchor Review Scope The fol lowing anchors are being assessed to verify that concrete expansion anchor factors of safety are in accordance with NRC I. E. Bulletin 79-02:

X-11 QC-1 X-11 QC-2 X-13A QC-1 X-13A QC-2 X-138 QC-1.

X-138 QC-2 X-168 QC-1 X-16A QC-2 X-23 QC-1 X-23 QC-2 X-24 QC-1 X-24 QC-2 X-47 QC-1 X-47 QC-2 X-7A X-91 X-7B X-9B X-7C X-10 Main Steam Frame - QC-1&2 X-7D X-17 X-8 X-12 Rec i rcul at ion Pi2e Re2lacement (RPRl Sco2e The anchors listed below were previously analyzed and qualified for QBE and SSE load cases.

In this assessment they are being analyzed for the pipe break load case only.

(Load combination 3}.

X-109A

~X-11 lA

.>>-X-1118

.:j. X-l 49A

~ X-1498 X-113 X-ll 6A X-1168 ISO Condenser Return Shutdown Supply Shutdown Supply Core Spray Core Spray Clean up Supply LPCI Pump Discharge LPCI Pump Discharge D-3 D-3 D-3 D-3 D-3 D-3 D-3*

D-3

COMB I NA TI ON L

DESCRIPTION I

OBE SSE PIPE BREAK NOTATION:

LOADING COMBINATIONS N

CONTRIBUTING LOADS AND LOAD FACTORS 0 ~

IA 1.ow + 1.OT + 1.OP + 1. OE0 + 1. 0 SRV lB

1. 41~ + l.7T + 1.7P + 1.7E0 + 1. 7 SRV

/lc-r

.1 'I Cf 1:.J {a;;l_.,,.

2 l.OW + 1.OT+ l.OP + 1 -OE 5 + LOCA 51~

ll fJ-'f"l l.OW + l.OT

+ l.OP + l.OR

,_3~

l.ow* + l.OT* + l.OP* + l.OR + 1.OEs + LOCA

~.J..

\\~ = Weight T = Pipe Thermal Reactions P = Reaction due to normal containment pressure (2 psig)

E0= Operating Basis Earthquake (OBE)

Es= Safe Shutdown Earthquake (SSE)

SRV = Reaction due to SRV Torus excitation of piping LOCA = Reaction due to LOCA Torus excitation of piping R = Pipe break loads (Use process pipe.Mp where moment 9iven is~~p)

W*= Wei9ht (Unbroken End}

TO BE USED 1*cR Elastic Analysis -

Anchor Frame Desiqn Elastic Analysis -

Rock Anchor & thru Bolt Assessment i

Elastic Analysis -

I All Components i11cl1Hl-j i ng Steel E111bc~d111cnts Elastic Analysis-All Components including Steel Embedments.

T*= Pipe Thermal Reaction (Unbroken End)

P*= Reaction due to containment pressurization accounting for timing with R.

32--

7_ l/

... -- \\;

~'

STRESS LIMIT COEFFICIENT ON APPLICABLE CODE APPLICABLE LOAD CASE LOAD CASE LOAD CASE ELEMENT CODE lA & lB (OBE) 2 (SSE) 3 (PIPE BREAK) l.6~.95Fy l.6~.95Fy STRUCTURAL STEEL MEMBERS AISC 1.0 (lA}

(Plastic Modulu~)

WELDS AISC l.O (lA) 1.6~.48 Fu(B) l. 6 ~. 48 Fu ( B )

y f... ~~

ACI 349 THROUGH-BOLTS

. 'i (~

APP. B l.O (lB)

l. 0

l. 0 Pd:>

ACI 349 WILLIAMS ROCK ANCHOR~ Tult APP.B l.O (lB)

1. 0

l. 0

~*

~i{

Pd<

ACI 349 U'.f~

r;Qth APP. B 1.0 (lB)

l. 0

1. 0 EXPANSION ANCHORS

}EDGE I. E. BULLETIN

.<._ '/

79-02

1. 0 (lA)

l. 0

l. 0 SHELL I. E. BULLETIN (5.. ~

79-02 l.O (lA)

l. 0

1. 0 ACI 349 SHEAR LUGS APP. B 1.0 (lB)

1. 0

l. 0 ACI 349 CONCRETE BEARING APP. B 1.0 (lB)

1. 0

1. 0

I 1Nteir41h,,,r;3

( j_ )

1HE EVfl.'....tJ.~-:-IGi~ OF Co~JC?.ETE ExPAr~sron ;._~<c:--ci=-s

[rTr::s STATION.

(/)

COMPLETION OF THE DRESDEN 3 RECIRCULATION PIPING REPLACEMENT (RPR) ANCHOR ASSESSMENT FOR FSAR COMMITMENTS.

(3)

COMPLETION OF ASSESSMENT OF THE QuAD CITIES CORE SPRAY f< ;/

Ut//oi,...5 FLUED HEAD ANCHOR FOR FSAR COMMITMENT, J:c.r4ccAVf-jJAf4 (4)

COMPLETION OF THE ASSESSMENT OF CONFIGURATION WALKDOWN

~

RESULTS FOR DRESDEN AND QUAD CITIES fLUED HEAD ANCHORS, (5)

A REVIEW OF THE NEED FOR FURTHER WALKDOWN EXPANSION BASED ON THE RESULTS OF ITEM 4, (6)

INCORPORATION OF ALL NEW NRR TECHNICAL COMMENTS (AS DISCUSSED IN OUR JANUARY 28~ 1988 EXIT MEETING) IN NEW CALCULATIONS WHICH ARE BEING PERFORMED AS PART OF THIS FLUED HEAD ANC~OR PROGRAM,

~ *--~*'

~.

Note The concrete cone capacity for the individual anchors and/or anchor assemblies shall be calculated as follows:

a. Concrete cone capacity Pd = 4.,_ ~ y fc x (Ap) where:

~ = 0.85 - All embedments are located in a compression zone or in a tension zone where the t~p.s..:j.Qn stress at the surface of the concrete is less than 5 ~ ~ fc.

ft = Actual in-place concrete strength-based upon concrete cylinder data.

Effective embedment depth will be determined in accordance with Fig. B7-l of ACI 349 Appendix B.

The effective projected net area (Ap = projected concrete cone area) shall be reduced to account for overlapping cones from adjacent anchors and/or assemblies as well as the in-fluence of the penetration sleeve.which intersects the cone surface.

The effect of the near-face reinforcing may be utilized t6 increase the capacity of the concrete cones.

The shear lugs alone shall be assumed to transfer the applied shear in accordance with the requirements of Appendix B since they are located in a compression zone.

b. The design strength for an individual anchor, where the cone capacity exceeds the ultimate tensile capacity of the rock anchor, shall be established as:

0. 9 As fy (As fut) shall be taken as 75,000 lbs. for l" diameter and 38,000 lbs for 3/4" diameter, Super High Tensile 'Spin-Lock" bolts in accordance with William's published data.

(As fy) shall be taken as 60,000 lbs. for l" diameter and 30,000 lbs for 3/4" diameter, Super High Tensile "Spin-Lock" bolts in accordance William's published design data *

c. For anchor assemblies in which the concrete cone capacity, Pd, is less than the ultimate strength of the concrete anchor, the design strength of the anchor shall not exceed:

.33 Pd for seismic load cases

• 5 Pd for pipe break load case Note See the Project Structural Design Criteria for ultimate capacities for wedge and shell anchors (DC-SE-01-CE).

The factor of safety is applied to the entire assembly rather than to individual anchors.

Shear and tension are combined using an elliptical interaction equation (5/3 power) *