ML19305C425
| ML19305C425 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 03/24/1980 |
| From: | Herring K, Persinko D Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19305C422 | List: |
| References | |
| TAC-11299, TAC-13152, NUDOCS 8003280532 | |
| Download: ML19305C425 (38) | |
Text
_
C 03/24/80 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of
)
)
Docket No. 50-344 PORTLAND GENERAL ELECTRIC COMPANY, ET AL. ) (Control Building)
)
(Trojan Nuclear Plant)
)
NRC STAFF TESTIMONY OF KENNETH S. HERRING AND DREW PERSINKO ON THE STRUCTURAL ADEQUACY OF THE
_ PROPOSED MODIFICATIONS TO THE TROJAN CONTROL BUILDING Q.1 Please state your name and position with the NRC.
A.1 My name is Kenneth S. Herring.
I am a Senior Structural Engineer in the Engineering Branch of the Division of Operating Reactors, Office of Nuclear Reactor Regulation.
A.1 My name is Drew Persinko.
I am a Structural Engineer in the Engineer-ing Branch of the Division of Operating Reactors, Office of Nuclear Reactor Regulation.
Q.2 Have you prepared statements of professional qualifications?
^
9 A.2 Yes.
Copies of our statements are attached to "NRC Staff Testimony of Kennuth S. Herring and Drew Persinko on CFSP Contentions 20, 12/13 and 16" which was filed on March 17, 1980.
Q.3 What are your responsibilities with regard to the NRC Staff's review of
~
the proposed modifications to the Trojan Control Building?
A.3 (Mr. Herring) As a Senior Structural Engineer, I have prime responsi-bility for the Staff's structural and mechanical review and evaluation 8003280 e
, of the proposed modifications. This includes a review to determine what structural effects actual modification work itself might have on existing structures as well as a review of the modifications to determine _,
whether they will substantially restore seismic margins to the Control Building Complex and bring that Complex into substantial compliance with the requirements of the Trojan license.
It also includes assuring that any effects of the modifications on the response of safety related systems, piping, equipment and components are adequately accounted for.
A.3 (Mr. Persinko) As a Structural Engineer in the Engineering Branch, I am responsible for assisting Mr. Herring in the structural review and evaluation of the Control Building modifications described by Mr.
Herring.
Q.4 What is the purpose of this testimony?
A.4 The purpose of this testimony is to present the Staff's basis for requir-ing modifications to the Trojan Control, Auxiliary and Fuel Building Complex (Complex), the Staff's position on the time dependence of interim operation, and the Staff's assessment of the structural adequacy of the proposed modifications to the Complex to restore the seismic margins to the Complex and to bring the Complex into substantial compliance with the requirements of the Trojan license.
In so doing, we will describe the unresolved items identified in the Staff's Safety Evaluation Report (SER) on the proposed modifications and indicate the status of resolution of those items.
, This testimony is also intended to provide responses to the struc-tural questions set forth by the Licensing Board at the Prehearing Conference on March 11, 1980, to address the structural aspects of Coalition for Safe Power (CFSP) Contention 22 and to address the remaining unresolvel. items with regard to the structural aspects of actual performance of the modification work itself.
I.
REASONS FOR REQUIRING MODIFICATIONS Q.5 Are the modifications required by the May 26, 1978 Order for Modi-fication of License considered applicable to the SSE as well as the OBE7 A.5 (Mr. Herring) Yes. The Order required substantial restoration of original design margins for the SSE as well as the OBE.
Q.6 Why is restoration of margins necessary?
A.6 (Mr. Herring) These margins are relied upon by the NRC in assessing the designs of older plants in light of current-day s&ismic design requirements.
Q.7 How have seismic design requirements changed since the time Trojan was licensed?
A.7 (Mr. Herring) A chronology of basic seismic design requirements, including those from around the time Trojan was designed to the present, is set forth below.
l l
i I
, The basic seismic design requirements have undergone many changes over approximately the past 25 years. Prior to 1960, there were no specific requirements other than those contained in local building codes. Since that time, the development of the basic seismic design practices can be generally summarized as follows:
PRIOR TO 1960 Uniform Building Code Requirements Static seismic coefficient applied to structures Ground motion described by Housner's 1960 - 1964 averaged ground response spectra.
Single degree of freedom systems were used for the evaluation of seismic responses.
Horizontal and vertical earthquake responses were not combined.
1965 - 1967 Ground motion described by Housner's averaged ground responses spectra (in some cases Housner made revisions from the previous spectra).
Multi-modal two dimensional codels were used for the evaluation of seismic responses.
The response spectrum approach was used most often. Time history was used occasionally.
Damping values were taken as 0.5% for piping.
1% 1/2% for steel structures, and 4% -
7-1/2% for concrete structures.
Compliance (flexibility) for plant foundation medium was considered.
Sum of the absolute value of the responses arising from the largest horizontal and the vertical earthquake was generally used for response determination.
l l
i
. 1967 - 1971 Ground motion described by Housner's averaged ground response spectra modified, especially in short periods, using Newmark criteria (known as modified Newmark spectra, 1967 - 1969).
~
Soil structure interaction effects were con-sidered using discrete soil springs and in some cases assuming material damping.
Floor response spectra generated and used in the evaluation of equipment and piping.
1971 - 1973 Modal damping values for the soil-structure system to represent contributions from both material and radiation damping limited to 10% of critical damping.
1973 - 1977 Reg. Guides 1.60 and 1.61 were introduced to define ground response spectra, and damping values (for structures, piping, equipment and components), respectively.
Damping for small and large piping was raised to 2% and 3%, respectively.
Soil damping determinations were required to account for the nonlinear stress - strain relationships for the foundation'edium.
m Finite element procedures were required in the calculation of soil-structure interaction for deeply em' bedded structures.
Three components of earthquake motion were
)
required to be considered by taking the SRSS l
of the responses to each component (Reg. Guide 1.92).
Floor response spectra generated per Reg. Guide 1.122.
AFTER 1977 Layered soils accounted for in an elastic half space soil-structure interaction analyses.
The limit of 10% of critical damping on modal j
damping values in soil-structure interaction analyses was removed.
-S-Comparison of elastic half-space and finite element soil-structure interaction analyses results.
Q.8 How do current design requirements compare to those used for plants such as Trojan?
A.8 (Mr. Herring)
In many respects they are more stringent, especially with regard to the definition of seismic loads.
Q.9 Why are older plants such as Trojan not required to be backfitted to meet current requirements?
A.9 (Mr. Herring) There are conservatisms in the design of the struc-tures of older plants, such as Trojan, which are relied upot in.
determining that backfitting to meet current requirements is not necessary.
In general, these conservatisms can be summarized as follows.
Conservatisms associated with the methodologies for seir ic analysis and design.
a.
Conservatisms for structures, systems, and components.
1.
Dynamic analysis.
Elastic dynamic analysis are performed using low damping values and time-history or response spectrum analysis methods.
In modal response spectrum analysis, closely spaced modes are combined by absolute summation.
2.
Soil sited structures evaluation.
Soil site structures are evaluated using conservative seismic inputs into soil-structure interaction analyses.
o 3.
Three input components.
Three input components of an earthquake (2 horizontal and 1 vertical) are considered. Both horizontal earthquake components are assumed to be equal.
'4.
Loading combinationa.
Loading combinations consider other loadings (e.g.,
dead weight, live loads, pressure loads, etc.) in addition to the seismic loadings.
Seismic loading is only a part of the total loading and in fact, other loadings besides seismic may in cases govern design.
b.
Effect of inelastic behavior.
In reality, well engineered structures, components and systems are capable of sustaining loads which are beyond those which would bring them to their elastic limit with-out sustaining damage. For small excursions into the inelastic range, seismic inertial loads arc reduced as a function of the amount of inelastic action in comparison with those calculated elastically. This phenomenon can be considered by the use of a ductility factor which is equal to unity for purely elastic behavior and increases with increasing inelastic behavior.
For example, a ducti-lity of 1.5 would have the effect of reducing accelerations of elastically calculated response spectra by as much as 1/3. Here ductility is defined as the ratio of displacement level in the nonlincar range to the displacement associated with the yield point for an elastic / perfectly plastic re-sistance vs. displacement function.
Conservatisms in-the structural and mechanical resistance.
a.
Allowable stress limits.
Engineering codes specify " code minimum strength" for materials.
These code minimum strengths are in turn specified by the applicant when the materials are ordered; any material found to be under that strength is rejected. The result is that the material supplier provides material of higher strength. Also, margins exist between allowable stresses and ultimate strengths.
b.
28-day concrete strength (structural only).
Designs are usually based upon the 28-day design strength of concrete. Concrete continues to gain strength with increasing time beyond 28 days. Additionally, the strength at 28 days often exceeds the called-for design strength.
c.
Static strength vs. dynamic resistance.
Code material strength, are based upon static load tests.
Since dynamic loads c antain a limited amount of energy and are applied at a faster rate, the margin between stress limits and failure for dynamic loads is greater than that for static
~
loads.
d.
Standard size structural members and pipes.
The design of the structural elements is such that their capacities usually exceed the requirements called for by the analyses. Much of the actual structural design is controlled by the availability of standard structural members such as beams and piping sections, so that larger sizes than are needed are often used.
e.
Redundancy in indeterminite structures and components allows for redistribution of loads.
From the standpoint of function, major structures and com-ponents can tolerate much deformation, and typically failure of numerous structural members. This deformation and loss of structural members can be sustained because of redundancy, (i.e., more char one path available to carry loads) which allows for redistribution of loads formerly carried by failed members.
f.
Ductility to failure.
In deforming to failure, beyond the elastic limit, the in-elastic behavior of well engineered coacrete and steel struc-tures, components and systems provides for energy absorption normally counted on in design.
g.
Minor attachments absorb energy.
Nonstructural elements which are not considered to carry any loads in design, do absorb energy through inelastic behavior or collapse during a seismic event.
Q.10 Does damping increase with increased nonlinear behavior of the structure?
A.10 (Mr. Herring) Yes, in the sense that " damping" is used to refer to increased energy absorbtion of the structure with increased nonlinear behavior.
_9_
Q.11 Can that increased " damping" be relied upon in determining con-formance with the Trojan design criteria?
A.11 (Mr. Herring)
No.
This increase in " damping" (or energy absorp-tion) is one of the items relied upon by the NRC in determining that it is not necessary to backfit the older plants to current seismic design requirements which have become more stringent with the evolution of NRC requirements. Also, it must be recognized that, on the basis of the results of the test program carried out by PGE in support of the proposed modifications, the proposed modifications will result in less conservatism inherent in the modified Complex than that which would have been present had there been no design deficiencies.
Q.12 Why are design rather than as-built material strengths now being used for capacity determinations for the modified Complex?
A.12 (Mr. Herring) These too are some of the conservatisms previously described as being relied upon by the NRC in determining that back-fitting to current design requirements is not necessary.
Q.13 Your testimony for interim' operation indicated that the structure was capable of resisting earthquakes in excess of 0.25g and, in fact, as high as 0.35g.
What was the basis for your judgment in this regard?
A.13 (Mr. Herring) - That judgment was based upon a 0.35g earthquake as defined by the Trojan FSAR seismic input criteria and an assessment y
v
. of capacity based upon extrapolition of test data in existance at
.the time which was not directly applicable to Trojan.
It was also based upon allowing for energy absorption through inelastic be-havior of the structure. Only the structure was considered. While the structures and systems within the Complex are felt to be capable of resisting earthquakes in excess of 0.25g as defined by the Trojan criteria, at some level below 0.35g there may be local failures of piping and equipment supports which were not factored into this consideration, and the type and extent of these potential failures were not analyzed.
Q.14 Does the Complex in its present configuration have appropriate margins to substantially meet the FSAR commitments?
A.14 (Mr. Herring)
No.
Q.15 Are the margins which are present adequate to provide for operation of the facility for the remaining duration of its operating license?
i A.15 (Mr. Herrfag)
No.
Q.16 Explain why they are not.
4 A.16 (Mr. Herring) As I discussed at the December 28, 1979 hearing session, it was deter 31ned to be acceptable for the facility to operate at reduced margin until appropriate modifications could be made to substantially restore the margins suggested by the initial
. design criteria. The time period necessary to implement the modi-fications, and that during which the margins would be reduced, is substantially shorter than the time remaining until expiration of the Trojan operating license. The concept of overall risk, as is ingrained in load combinations, provides the basis for this.
Q.17 What is the time for which interim operation shoald be allowed?
A.17 (Mr. Herring) As I discussed at the December 28, 1979 hearing session, there is no explicit time limit although the length of interim operation is time dependent. Operation for a period on the order of about 3 to 4 years or so from the issuance of the May 26, 1978 Order would be appropriate.
II.
STRUCTURAL ADEQUACY OF THE PROPOSED MODIFICATIONS Unresolved Items in The Staff's SER of February 14, 1980 Q.18 Please identify and describe the significance of the unresolved items with regard to structural adequacy of the proposed modifi-cations that are listed in the SER.
A.18 (Mr. Herring) The unresolved items which have a bearing on determin-ing the structural adequacy of the proposed modifications, as identified in the SER, are:
(1) Method of accounting for the encased steel frame in deriving stiffnesses (SER 95.1.1.1, p.63).
In this regard, the effect of double curvature behavior was not accounted for and it was not shown that double curvature behavior would not occur. Moreover, the j
licensee's assumption that slip in the beam-column
3 i
1 i connections would be sufficient to develop twice the AISC allowables was not demonstrated as valid. The significance of this is that the stiffness derivation for the modified Complex has not been shown to be ade-quate insofar as stiffness is dependent upon proper treatment of the encased steel frame.
(2) Dead Load Determination (SER 85.1.1.3, p.65).
Stiff-ness of the structural elements is proportional to the normal forces on the elements and the normal forces are dependent upon the dead load and the vertical earth-quake components.
In this regard:
(a) The effect on dead load of creep and shrinkage was not adequately quantified; (b) The assumed value of shrinkage strain was not adequcte2y considered; (c) Stiffening of beams due to encasement in concrete and the effect of this on dead load was not properly considered; and (d) The effect of a 50*F change in mean temperature on dead load reduction for exterior walls was not addressed.
Each of these matters will affect the dead load in walls.
Without properly determined dead load, normal forces and, therefore, stiffnesses cannot be correctly determined.
(3) Gross Bending Moment Effects on Stiffness (SER 95.1.1.3, pp. 66, 68). Gross bending moments from an earthquake will cause shifting in wall normal forces and, therefore, shifting in stiffnesses. Any tension induced in walls from this gross bending moment effect must be shown not to be detrimental to stiffness over a number of cycles (the licensee's test program accounted only for com-pression, not for tension effects).
The effect on stiff-ness of tension and cycles of tension from the gross I
bending moments must be quantified before stiffness of the modified Complex can be adequately known.
. (4) Single vs. Double Curvature Mode of Failure (SER 85.1.1.4, p.68).
The licensee's test program did not demonstrate the actual behavior of walls in the Complex and whether the single or double curvature mode of failure would occur.
Stiffness is dependent upon whether walls behave in the single or double curvature mode. Consequently, because neither mode of behavior has been fully demonstrated te be applicable to the Complex exclusive of the other, both modes of behavior and their effects on stiffness should be accounted for.
(5) Capacities of new structural elements (walls and plate) -
slippage and the coefficient of friction between steel and concrete (SER 95.2.1, p.69). Stiffness in the structure will decrease due to overturning moments and single curvature behavior. The new structural elements =ust be capable of withstanding these effects and this, in turn, is dependent upon slippage and frictional resistance.
The use of a 0.7 coefficient of friction between the steel plate and concrete,-relied upon to transmit seismic forces to the plate, requires justification. Similarly, the resistance to sliding between columns and footings must be justified.
The capacities of the new structural elements, and therefore of the modified Complex, is dependent upon a demonstration that the slippage and friction assumptions made are justified.
(6) Capacities relied upon can be developed (SER 85.2.2.1, p.71).
Each wall panel in the Complex must be capable
- of carrying the forces calculated and relied upon to be withstood for the flexure, sliding and diagonal tension (shear) modes of failure.
This must be verified for each element of the modified Complex before a conclusion on the capacity of the modified Complex may be reached.
. (7) Flexure Mode of Failure and Flexural Capacities (SER 85.2.2.1, pp. 72-73).
For a proper determination of flexural loads and capacities, the following items must be resolved:
(a) Dead load contribution to normal forces will affect flexure loads and capacities. Conse-quently, those unresolved items delineated in items (2) and (3) as affecting dead load must be resolved in order that the proper dead load contribution to normal forces and the effects on flexure loads and capacities can be determined; (b) Dead load effect on flexure capacity of indi-vidual wall panels. While the dead load contributes only 6% of overall flexure capacity for the entire structure and, thus, the un-resolved items with regard to dead load should have little effect on the overall flexure capacity of the structure, the effects on flexure capacity of individual wall panels could be more significant.
Such effects on individual panels must be exn. mined; (c)
If single curvature behavior is assumed, certain displacements must take place in order to develop the necessary resistance to flexure failure.
It must be shown that these displacements can take place and that they are compatible with the defor-mations of the structure.
In addition, if the required displacements do occur, the resulting vertical shear forces at some places on the R and N walls may exceed capacities. The acceptability of this must be demonstrated.
(8) Sliding Mode of failure and sliding capacities (SER 85.2.2.1, p.73).
Shear friction contributes to the resistance to sliding failure. The licensee's formulation of the shear friction re-sistance to sliding failure is inadequate and gives too large a resistance for the Trojan walls. An appropriate relation-ship for the shear friction resistance to sliding must be used before a correct determination of capacity against sliding can l
be made. Also, this resistance mechanism is affected by the l
matters discussed in items (2) and (3) above.
l l
l
. (9) Displacements as affected by stiffness, frictional resistance to sliding and gross overturning moment effects (SER 85.2.3, p.74).
The elastic displacements determined from the STARDYNE model may be increased by stiffness degradation, frictional
~
resistance to sliding and gross overturning moment effects.
Thus, the following unresolved matters were identified as affecting final calculated displacements:
(a) Stiffness effects. Displacement depends on stiff-ness but unresolved matters remain with regard to stiffness derivations as indicated under items (1),
(2), (3) and (4).
(b) Shear friction.
Shear friction resistance to slid-ing will affect displacements and the faadequacy in the shear friction resistance formulation described in item (8) must be corrected.
(c) Cross overturning moments. Gross overturning moments affect displacements but such effects were not addressed.
(10) Floor response spectra as affected by stiffness (SER 95.3, pp. 74, 75).
The floor response spectra for the modified Complex is dependent upon stiffness and stiffness degradation.
Before final floor response spectra can be properly derived, the unresolved matters regarding stiffness as described in items (1), (2), (3) and (4) must be satisfactorily resolved.
(11) Cyclic effects (SER 95.5).
The cyclic effects of the occur-rence of multiple earthquakes will degrade the stiffness of the structure. The unresolved items with regard to stiffness described in items (1), (2), (3) and (4) must be resolved and accounted for before the ability of the modified Complex to withstand multiple earthquakes can be finally determined.
(12) Shrinkage increase with decreasing wall thickness (SER 95.6, p.76).
Shrinkage increases as wall thickness decreases. This phenomenon should be addressed to assure that shrinkage, which is important because of the encased steel frame and which has substantial effects on dead load, has been adequately accounted for.
i
. (13) Final capacity to force ratios and wall degradation (SER 55.12, p.83).
The unresolved matters described in items (1)-
(8) and (12) will affect the determination of capacities and
~
forces for walls in the Complex. Final capacity to force ratios are needed before the potential effects of wall degra-dation for walls with capacity to force ratios less than one, and the effects of such degradation on equipment in the vicinity of those walls can be finally determined.
Q.19 As to unresolved item 2(a), provide the bases for your determination,
-6 in as expressed in the SER, that up to 140 x 10 M is a appropr-late value of restrained shrinkage strain to be use in calculating the maximum dead load reductions to be expected for the existing walls.
A.19 Shrinkage is a volume change of concrete and is an inelastic defor-mation that is caused by a loss of water as curing progresses.
Ic is a complicated phenomenon that is independent of externally applied loads and temperature imposed changes. The American Concrete Insti-tute (ACI) suggests a method of calculating unrestrained shrinkage strain in its publication No. SP 27-13.
Here, the unrestrained shrinkage strain is a function of ultimate shrinkage strain, time, humidity, member thickness, slump, cement cont'ent, percent fines and air content. This unrestrained value will be modified by any restraints in the actual situation, such as rebars. A method for performing this calculation is presented by Park and Paulay in their book " Reinforced Concrete Structures" where restrained shrinkage strain is calculated for a section of concrete restrained by rebars.
Both of the above references deal with only reinforced masonry wall.
The licensee has utilized the approach suggested by the above L_
. references. However, complications arise due to the introduction of the masonry aythes which sandwich the concrete core. The above references do not address specifically this situation and therefore judgments must be made.
In the licensee's calculation of restrained shrinkage, the masonry wythes were counted in the overall wall thickness similar to concrete and also as restraint similar to rebar. The thicker a vall is, the lower the shrinkage strain will be at a given time.
It appears appropriate to count the masonry in ' btaining wall thick-o ness because it will obstruct the flow of moisture to the atmosphere which will be at the face of the masonry and thus lessen shrinkage.
However, the pre-shrunk =asonry blocks will expand as they contact the fresh concrete in the core and the water begins to flow through the masonry to the air surface. Shrinkage is a reversible phenomenon.
The masonry block will then shrink again as moisture leaves the wall system at the block air interface. Thus, the block behaves differently than the rebar and any restraintit offers is difficult to estimate considering long term effects.
Taking the 70 x 10~0 I"/in as calculated by the licensee and count-ing the macor.ry in overall wall thickness but not as restraint, one obtains 123 x 10-6 in/in. Taking a two dimensional effect into account through Poissons ratio of u =.15 yields a restrained strain of as much as 141 x 10-6 in/in. This value is for a 30" thick wall and will increase as wall thickness decreases.
In addition, a Poisson's ratio of 0.21 was previously indicated by PGE as being appropriate for the in-situ walls.
i Thus a value of restrained shrinkage strain of 140 x 10-6 injf,
appears to be a reasonably conservative value. Additionally, although even discontinuous core steel will provide restraint to shrinkage above that relied on above, substantial reliance cannot be placed on the existence of core steel in the composite walls throughout the Complex to resist shrinkage, although some may be discontinuous. The March 20, 1980 letter from PGE to the NRC regard-ing reinforcing steel in the Complex shear walls indicates that no composite wall panels in the Fuel Building contain any (vertical or horizontal) reinforcing steel in the concrete core, few (about 15%)
of the composite wall panels in the Auxiliary Building contain any (vertical or horizontal) reinforcing steel in the concrete core, and only about 60% of the panels above Elevation 93' and about 94%
of the panels below Elevation 93 in the Control Building contain any (vertical or horizontal) core reinforcing steel.
Q.20 What is the status of resolution of the unresolved matters raised by the Staff?
A.20 The licensee has made an attempt to resolve the matters raised by the Staff and identified as unresolved items in the SER dated
~~
February 14, 1980. The NRC staff me'. with PGE and Bechtel on March 7 an'd 8, 1980. Following this meeting, the licensee submitted addi-tional information in a letter dated March 17, 1980.
Also, the licensee addressed unresolved items in its pre-filed testimony of March 17, 1980.
I l
r
These documents contain substantial additional infor=ation which should be of significant valuu in proceeding with resolution of the items. However, it should be noted that licecaee's letter of March 17 was received late in the afternoon on March 18.
Pre-filed testimony was not available for staff review until March 19.
In addition, some information needed for staff testimony (licensee's letters of March 20 and 21, 1980) was received on March 21 and 24, 1980.
In consideration of the bulk, substance and timing of these filings
-- some of which contain new information -- the staff has not had an adequate amount of time to give a considered review to these important matters.
The Staff as been engaged in an intensive review of licensee material since its receipt. Nevertheless, although progress has been made, the unresolved items raised by the Staff in the SER remain unresolved at this time. The NRC Staff review will continue on a first-priority basis until the start of the evidentiary hearing commencing March 31, 1980.
LICENSING BOARD QUESTIONS Q.21 At the Prehearing Conference held in this proceeding on March 11, 1980, the Licensing Board set forth a number of questions bearing on the structural adequacy of the modified Complex. With' regard to the criteria for determining whether the proposed modifications will substantially restore the seismic margins and bring the Control
. Building into substantial compliance with the Trojan license, the Board asked:
(1) What are the criteria that we should use to assure that the Control Building is brought into substantial com-pliance and the intended margins met?
(Tr. 3531).
(2) On what basis will it be determined that the modified structure will have increased seismic capacity to safely resist the 0.15g OBE forces with the margins inherent in the original design criteria?
(Tr. 3531-32).
(3) How do you assure yourself that you have met the original design criteria and are in substantial compliance with that criteria as set out in the technical specifications?
(Tr. 3532).
Please respond to these Licensing Board questions.
A.21 The basic seismic design requirements for the Complex have been set forth in-Section 3 of the Staff's SER.
This Section references the appropriate portions of the Trojan FSAR, as referenced by Trojan Technical Specification 5.7.1, and discusses the degree to which they are met, as determined by the NRC Staff review. Rather than demonstrating substantial literal compliance with all appropriate design requirements, the results of a testing program were imple-mented by the licensee to demonstrate the capability of the in-situ walls. Given the required seismic input definition, it must be-demonstrated that sufficient margin exists in the modified Complex, including the new and existing walls, to resist the loads resulting from the use of these inputs, and that uncertainties over the actual behavior of the structure when extrapolating the res:1.ts of the testing program to the behavior of the in-situ walls are ade-quately accounted for. These uncertainties arise from the effects
~ ~
. of the encased frame on lateral load resistance, the higher stress levels being present in the walls than would have resulted from appropriate design of the structure initially, and the sensitivity of the stiffnesses and capacities to the parameters contributing to them, as indicated by the testing program.
It should be noted that FSAR Section 3.8.1-5.1 specifically states that the structural steel framing for the Control, Auxiliary and Fuel Buildings was initially intended to carry only vertical loads, while the lateral loads due to earthquake, wind, and tornado are resisted by reinforced concrete and concretc block shear walls. However, the encased steel frame is now being relied upon to supply lateral load resistance. Since this element of conservatism which was present in the original design but was not relied upon is now being relied upon, it requires a careful assessment of the structural response and the capacity to force ratios for the walls.
Q.22 With regard to the original design of the Complex, the Licensing Board asked:
(9) How was the construction of the composite walls taken 4
into consideration in meeting the building code requirements for the original specifications and construction when there was no Uniform Building Code requirement appropriate for that kind of l
construe:4on? (Tr. 3533).
Please respond to_this. Licensing Board question.
A.22 As indicated in the PGE submittals regarding the proposed Control Building modifications, there were no explicit Code (ACI or UBC) l l
requirements for the composite (major shear walls) in the Complex.
l The initial design concept for these composite walls for in-place l
l
, loads neglected the area of the block and relied upon only the equivalent wall thickness, determined from considering the cell grout and concrete core, subject to the Ultimate Strength Design requirements of ACI 318-63. Out-of-plane wall capacities were based on ACI 318-63 Ultimate Strength Design concepts. The single wythe and mortared double wythe masonry block walls were designed using the load combinations for reinforced concrete, including the load factors, and the allowable stress was defined as an in-crease factor times the UBC allowable stresses. The details of the criteria aretsummarized in the response to NRC question 2 in PGE's December 31, 1979 submittal regarding the " wall problem".
Q.23 With regard to demonstrating the adequacy of the modified Complex under the building codes, the Licensing Board stated:
(4) We should know just how the building codes permit this kind of test results to be used in meeting the code specifications.
(Tr. 3532)
(12) Show whether the Uniform Building Code provides for the use of, and allows accounting for, the more sophisticated analyses of the seismic forces provided by the STARDYNE analysis.
(Tr. 3533)
Please provide your responses.
A.23 Sections 106 and 107 of UBC 1967 allow for the determination of
~
structural strengths based upon testing. ~his provision is also included in Section 104 of,ACI 318-63. Additionally, much of the design criteria within the codes are established through the evalu-ation of the results of testing programs.
Guidance is not given in either the UBC or the ACI Codes regarding acceptable methods of l
l u
. performing tests to establish design criteria for loads such as earthquake. Only static load tests of actual structures are explicitly addressed (see Section 24 of UBC 1967 and Chapter 2 of ACI 318-63).
While the ACI Code does allow for departure from certain design rules or formula on the basis of analysis, it is not felt that the STARDYNE analyses which have been,erformed for the determination of the loads in the various wall elements in the Complex is sufficient to qualify for reduction in allowable stresses for the thus determined loads.
The Codes do require that the loads in structural elements be determined using sound principals of engineering mechanics. Appropriate consideration must be given to the complexity of the structure when choosing a particular analytical technique for load determination.
Limitations on the chosen analytical technique should be recognized and adequately considered in determining the final load on a particular element for design purposes. Appropriate Code provisions should then be used to establish the design of the structural element.
It is felt that the STARDYNE analyses can give appropriate load definition for the walls in the Complex, given that appropriate stiffnesses for the structural elements are used and that any uncertainties which cannot be incorporated into the STARDYNE analyses are adequately accounted for in the evaluation and designs of the structural elements.
4
. Q.24 The Licensing Board made inquiries with regard to the testing pro-gram and results which are being used in the analysis of the modified Complex.
In this regard, the Board asked:
(10) Show how a 32-icch type A wall with a vertical load of 105 psi. would have a unit shear capacity 18 per-cent greater than a similar 24-inch wall.
(Tr. 3533).
Please respond.
A.24 The referenced example of an increase in unit shear capacity with increasing wall thickness which is the subject of this question was set forth in the initial version of PGE-1020, was deleted in Revision 1 to PGE-1020 and has not been reinstated in subsequent revisions to that document.
The current criteria for the walls involves an investigation of three modes of failure, namely flexural (both double & single curvature), sliding, and diagonal tension (shear) modes of failure.
The unit shear stress capacities for the flexural modes are inde-pendent of wall thickness. Therefore, there would be no difference in this capacity for the thicker walls.
The sliding mode capacity is governed by vertical reinforcement ratio, embedded column capacity, and a contribution from the normal force which is a function of the ratio of the area of concrete to the area of mortared block. The latter contribution to capacity would increase with increasing core (vall) ~:thi'ckness. The exact percentage increase depends on the magnitude of the other contrib-uting factors (i.e. vertical reinforcement ratio, embedded column capacity).
. l The diagonal tension capacity is a function of horizontal and
. vertical reinforcement ratios, the compressive stress on the wall, and the ratio of areas of block to areas of concrete cores. This capacity would increase with increasing core (wall) thickness, the exact percentage increase depending on the magnitude of the other contributing factors.
Each of these criteria is investigated and relied upon for each in-situ wall (see March 17, 1980 PGE submittal in response to Staff's March 7, 1980 requests for additional information). Therfore, this additional element of conservatism alluded to in the initial version of PGE-1020 is no longer present.
Q.25 (20) Determine if there is any reason to believe that there might be a scaling factor for the ability of walls to withstand seismic forces with the same aspect ratio but much larger as occurs in the Control Building compared to the test models.
(Tr. 3534-35).
A.25 The test specimens were of sufficient size such that when it is considered that the walls of the Complex meet the height to thick-ness ratios of the UBC, it is felt that the criteria that have been developed based on the results of these test specimens are adequate for the in-situ walls. No scaling factor is necessary.
1
s
. Q.26 Also with regard to the test program:
4 (21) Determine if there is a different effect on the stress versus the displacement in a wall if the frequencies of cycling were high as in an earthquake compared to the test frequencies used for the wall evaluations.
(Tr. 3535)
A.26 The effects of higher frequency cyclic loading should reduce the stiffness degradation relative to that demonstrated by the licensee's test program in which specimens were tested in a psuedo-static manner.
However, the effects of higher frequency loading cannot be quantified as they were not the subject of the licensee's test program.
Q.27 The Licensing Board made a number of inquiries with regard to the seismic analyses for the modified Complex and input to those analyses. Specifically (5) How do you conclude that the earthquake ground response spectra has a vertical ground acceleration that's two-thirds of the horizontal acceleration?' (Tr. 3532).
A.27 As stated in Section 3.2.1.1.1 of the Staff's SER, the design response spectra and peak ground acceleration are specified in Section 3.7.1.1 of the Trojan FSAR. As stated in the Staff's SER, these have been incorporated into the seismic analyses of the modified Complex. The referenced FSAR section defines the SSE and OBE response spectra.las well as the peak ground accelerations to be assumed for the OBE and SSE in the vertical direction, namely 0.lg vertically for the OBE l
(or 2/3 of the 0.15g horizontal acceleration), and 0.17g vertically for the SSE (or 2/3 of the 0.25g horizontal acceleration).
l l
, Accordingly, the vertical ground accelerations are two-thirds of the horizontal accelerations under the licensed criteria for the plant.
Q.28 (7) The existing wall capacities are determined based on the testing results using the total dead load on the wall reduced by 20 percent to account for the vertical earthquake effect. How and why was that done? (Tr. 3532).
A.28 The dead load is being reduced by 13%, not 20%. The basis for this under OBE conditions is that the vertical rigid range OBE acceler-ation is 2/3 (.15g) =.lg, amplified by 30% to account for vertical building response.
Thus, 1.3 x.lg =.13g or 13%g. This vertical motion fluctuates between 13%g, and is thus the basis for a 13%
dead load reduction to account for vertical earthquake motion.
If an amplification of 16% (claimed by the licensee to be more representative such that the use of 13% dead load reduction is
~
~
~
" conservative" relative to' this more representative value) is used, this is reduced to about 0.12g (1.16 x 0.lg) or 12%g, thus indicat-ing that the degree of " conservatism" is not substantial (i.e. 1%g).
Q.29 (11) Show how the structural steel columns in the shear walls will be used in determining the failure limitations of the Control Building walls, if any, or if it is used as a safety factor.
(Tr. 3533).
l A.29 The encased steel frame, which includes the steel columns and beams, l
1s relied upon to provide lateral resistance in the determination of l
~,
. capacities for single curvature flexure behavior and sliding.
In addition, the beam-column connections are relied upon to resist gross vertical shears along the column lines. Therefore, the columns and the beams are not being taken as added factors of.
conservatism. The encased framing, while contributing to the resistance against modes of failure other than double curvature, does not contribute to the double curvature capacity, nor to double curvature stiffness.
Q.30 (16) Provide an evaluation of the temperature coefficient of expansion effects that might take place between the steel place and the con-crete wall to which it is bound and tensioned once the concrete wall has reached full strength.
(Tr. 3534).
A.30 The effects of temperature changes and thermal expansion have been factored into the design of the steel plate. The effects of tempera-ture changes and thermal expansion have also been accounted for in
- the calculation of bolt tension losses, and-the analyses of the effect of the plate on wall capacity.
Q.31 (17) Show whether there are any tension effects in the bolts that in-fluence wall strength.
(Tr. 3534).
A.31 In the analysis of the effect of the steel plate on strength of the Control Building west wall reliance is placed on 75% of the initial bolt preload to resist shear forces in the building. The 25%
allowance was based uponnconsideration of the losses in bolt tension which will occur over time due to bolt relaxation, concrete creep I, __-
, and shrinkage and temperature effects. Accordingly, it is necessary to assure that, on the average, 75% of bolt preload is available and that this level of preload is uniform.
~
To provide such assurance an in-service inspection program should be required. After review of the licensee's initial proposal, and the modified proposal contained in the licensee's "non-structural" testimony of March 17, 1980, and consideration of the additional requirements contained in Section 6.1.2 of the SER of February 14, 1980 on this subject, the following in-service inspection program should be implemented:
(a) Demonstrating that each bolt in a random and represen-tative sample of not less than 25 percent of the total number of bolts has a tension equal to or greater than 80 percent of the initial bolt tension.
If the tension in any bolt is below 80 percent of the initial bolt ten-sion, the tension in two adjacent bolts shall be measured.
If either of these bolts is found to have
~
less than 80 percent of the initial bolt tension, then all bolts shall be tested.
All bolts found to have less than 80 percent of the initial bolt tension shall be retensioned to the original installation tension value.
(b) Demonstrating the acceptability of the test sample by showing that E-2a is greater than.8X, where i is the mean sample tension, o is the standard deviation and x9 is the mean initial bolt tension.
If this criterion is not met, then all bolts shall be tested to the criteria in (c) above.
If the mean sample tension i is below.8X,,
the circumstances shall be reported pur-suant to Technical Specification 6.9.1.8.1.
(c) Determining that there is no evidence of degradation or abnormal conditions by visual inspection of the condition
, of all bolts in the sample, their end anchorages and con-crete or masonry in the vicinity of the anchorage.
In addition, six (6) surveillance bolts, three (3) in the R-line wall and three (3) in the N-line wall, shall be
~
removed and inspected and the condition of the tape should be noted, as well as the condition of the bolt surface to ensure that the tape and bolt will continue to perform their functions with the design safety mar-gins present in the bolt. Abnormal degradation of the bolt shall be reported pursuant to Technical Specifi-cation 6.9.1.8.1.
(d) If the bolts inspected during the first four inspections meet the acceptance criteria of (a), (b) and (c), then the sample for the subsequent inspections may be reduced to not less than 10 percent of the total number of bolts.
(e) Beginning with the third year inspection, a trend analysis shall be used to predict the existing bolt tension at the end of the next inspection interval (2 years or 5 years) thereafter.
If the predicted bolt tension is below.75 of the initial bolt tension (x - 26 f. 75 X,), corrective action shall be taken in accordance with (a) above, and a detailed inspection report shall be submitted pursuant to Technical Specification 6.9.2 within 90 days after com-pletion of the surveillance testing.
O i
, Q.32 With regard to the rail stop to be installed in the Turbine Building railroad bay adjacent to the newly installed west wall of the Con-trol Building, the Licensing Board stated:
(18) Determine if there is any foreseeable possibility of an impact on the rail stop in the Turbine Building that might result in a force against the face of the Control Building if the stop were to fail and what this would do to the construction of the wall where it would im-pact the Control Building.
(Tr. 3534).
Please respond.
A.32 An impact of a train on the rail stop cannot be totally ruled out.
The consequences of hitting the rail stop is a function of both the weight and the speed of the train. Depending upon the neight of the train, the rail stop will prevent an impact on the Centrol Building west wall for a train traveling at very low speeds. At higher speeds and weights, the rail stop alone will not be adequate to stop a train and penetration of the west wall of the Control Building could occur. We have not performed an analysis to determine the precise combination of train speeds and weights that could result in a wall penetration nor have the consequences of such penetration been evaluated.
Rather, as set forth in Section 5.7 of the Staff's SER and in "NRC Staff Testimony of Charles M. Trammell, III on Questions Regarding Relocation of the Railroad Spur and Reduction in Size of An Equipment Hatch Under the Proposed Modifi-cations," filed on March 17, 1980, administrative controls should provide assurance that movement of trains on-site will be adequately
, controlled so as to preclude impacts that will result in damage to the Control Building.
CFSP CONTENTION 22 Q.33 CFPS Contention 22 states The effect of the steel place on displacement in the Complex has not been completely analyzed.
What displacements can occur during an earthquake?
A.33 An earthquake will cause displacements of structures and components.
These diplacements are time dependent and at any particular time during an earthquake, different parts of structures may undergo different displacements in different directions. From the stand-point of the effects of the steel plate, the displacements of concern are interstructure displacements - that is, relative dis-placements between two adjacent structures, in this case, between the Control Building and the Turbine Building.
Q.34 Why are such relative displacements a concern?
A.34 Depending on the magnitude and the direction of the ddsplacements for each building, building contact could occur. For interim operation, an analysis was performed which demonstrated that the maximum relative displacements of structures caused by an earthquate were less than the gaps between structures, with margin, at all elevations, thereby assuring that building contact would not occur. The concern with installation of the steel plate on the west wall of the Control t
i I
l
, Building under the modifications is that the three-inch thick plate will reduce the gap available between the Control and Turbine
]
Buildings at elevations 93' and 69' where Turbine Building girders and floor slabs are located.
Q.35 What measures will be taken to compensate for the reduced gaps where the steel plate will be installed?
A.35 At elevation 93', three inches will be removed from the flange of the steel girder in order to increase the gaps between the installed steel plate and the slab and girder. At elevation 69', 18 inches of the overhanging part of a concrete slab will be removed. These modifications will provide a gap of about four inches for displace-ments in the north-south (N-S) direction (parallel to the steel plate) and gaps in the east-west (E-W) direction (perpendicular to the plane of the steel plate) of two inches at elevation 93' and 2.5 inenes at elevation 69'.
Q.36 What effect will the steel plate and the added walls from the modi-fications have on displacement of structures?
A.36 The steel plate and added walls should have no effect at all on displacement of the Turbine Building. The plate and added walls will stiffen the Control Building Complex in the N-S direction and will thus cause displacements in that direction to be less for the modified Complex than for the as-built Complex. The addition of walls and the plate will not significantly stiffen the Complex in the E-W direction, however, and, therefore, displacements in that l
=
m.
- direction will not be significantly reduced relative to those for the unmodified structures. Also, there would not be a significant reduction in the E-W displacements of the Complex due to the pro-posed " structural improvements" to the walls running parallel to the E-W direction.
Q.37 What assurance is there that the material removal that you described will prevent contact between the Control and Turbine Buildings?
A.37 It was shown in Phase I of this proceeding that for N-S displace-ments, a gap of three inches between buildings is sufficient to preclude contact of the as-built buildings during an earthquake.
Since displacement of the Control Building in the N-S direction will be reduced by the addition of the three walls and the steel plate, a gap of four inches after the modifications should be adequate.
It was shown in Phase I of this proceeding that for E-W displacements, a gap of two inches at elevation 99' was adequate to maintain clearance between the buildings with margin. While dis-placements in the E-W direction will not be significantly reduced by the modifications, gaps of two inches at elevation 93' and 2.5 inches at elevation 69' after the modifications should be adequate to prevent building contact. The precise margins against contact of the buildings however can only be fully quantified when the unresolved matters affecting displacement (identified in item 9 in response to question 2 of this testimony) have been resolved.
m*
, Conclusion on Structural Adecuacy of the Proposed Modifications Q.38 Based on the analyses and evaluations that have been performed and on the status of the unresolved matters previously identified, what is your conclusion with regard to the adequacy of the proposed modifications to substantially restore the seismic margins and bring the Control Building into substantial compliance with the requirements of the Trojan license?
A.38 Although some progress has been made in completing our review, the unresolved items identified in the SER are not fully resolved, as discussed in the answer to Question 20 above. Since the Staff's ultimate conclusion regarding the adequacy of the proposed modifi-j cations hinges on the final resolution of these items, we have not yet reached a final conclusion.
III. SEISMIC CAPABILITY OF THE COMPLEX DURING PERFORMANCE OF THE MODIFICATION WORK Q.39 In the Staff's SER of February 14, 1980, several unresolved items with regard to structural aspects of actual performance of the i
modification work were identified. Please identify and describe the significance of those unresolved matters.
i A.39 The unresolved matters having a bearing on the structural adequacy of performance of the modification work itself were:
(1) A-frames for transport of plate 8 (SER S4.6, p.39, 84.11.1, p.54).
A-frames will be attached to plate 8 for transporting it across the Turbine Building operating floor and to prevent a flat drop of plate 8 onto the floor. At the time of preparation of the
, SER, the A-frame design had not been finalized.
Con-sequently, its adequacy to prevent a flat plate drop and therefore to prevent damage to the operating floor from such a drop could not be determined.
(2) Jacking plate 8 above the Turbine Building operating floor (SER 94.11.1, p.54).
Potential problems from jacking the edge of plate 8 have been eliminated since the plate will not be jacked up during handling.
(2)
Exposing columns to tie-in reinforcing steel (SER 94.11.3, p.56).
Certain columns will be exposed at various elevations to tie-in reinforcing steel from new walls and to make existing reinforcing steel i
continuous. Simultaneously exposing a number of columns can reduce the capacity of the structure significantly. Accordingly, it is necessary to show that simultaneous opening of columns will not signifi-cantly reduce seismic capacity or that this work will be done in a sequence such that adequate seismic capacity is maintained.
(4) Seismic qualification of safety-related equipment, components and piping during modification work (SER 84.13, pp. 57-58).
Since the seismic qualification of equipment and the like during the modification work is dependent, in part, upon the final floor response spectra for the modified Complex, the unresolved items with regard to such final flocr response spectra (see item (10) in response to Q.UB in this testimony) must be resolved before the proper modifications to equip-ment, components and piping to assure qualification during performance of the modification work can be made.
,-___m
m'
. Q.40 What is the status of resolution of these matters?
A.40 With regard to the adequacy of the A-frames, while the licensee's design of the A-frames has been completed, details of design criteria and load factors were not received until March 24, 1980. Detailed load combinations and corresponding acceptance criteria must, at a minimum, be provided in order for the Staff'c review and determi-nation of the adequacy of the A-frames to be completed.
Similarly, we have not completed our review of the new information on the sequence and timing of opening columns (item (3)) to determine whether adequate seismic capacity will be matinained during this work.
Finally, as to the seismic qualification of safety-related equip-ment, components and piping during the modification work, this is dependent upon the derivation of appropriate floor response spectra w
for the modified Complex. As previously indicatet. unresolved questions remain as to these floor response spectra and Staff work and review is continuing on these matters in an effort to reach resolution.
Q.41 What is your conclusion with regard to structural aspects of the modification work itself?
A.41 Until the unresolved matters discussed in response to question 39 and 40 are resolved, we are unable to conclude that all structural aspects l
of the modification work itself have been resolved and that adequate
e<L O
, seismic capability will be maintained throughout the performance of the modification work.
?
s f
t h
I I
f
,