ML20101L677

Plant Vogtle Settlement Review

ML20101L677
Person / Time
Site:	Vogtle
Issue date:	09/30/1984
From:	GEORGIA POWER CO.
To:
Shared Package
ML20101L675	List: ML20101L671 ML20101L677
References
NUDOCS 8501020332
Download: ML20101L677 (47)

Text

'

.)

<4 t

l t

1 PLANT VOGTLE SETTLEMENT REVIEW SEPTEMBER, 1984 i

I Appendix A revised 12-19-84 f

8501020332 841229 PDR ADOCK 05000424 E PDR

j a ..-

Contents

1. INTRODUCTION AND

SUMMARY

2. ESTIMATED SETILEMENT 2.1 PSAR Estimated Settlements 2.2 FSAR Estimated Settlements

3. REVIEW 0F MEASURED SETILEMENT

4. REVISED SETTLEMENT ESTIMATE

5. CONCLUSIONS

6. REFERENCES Tables 1-1 Estimated and Recorded Settlement 3-1 Summary of Recorded Settlement 3-2 Differential Settlement Setween Structures Ficures l 3-1 Turbine Building Settlement 3-2 Control Building Settlement 3-3 Fuel and Auxiliary Building Settlement 4-1 Lower Sand Modulus 4-2 Revised Settlement Eage 2

~

.lf

. v .-

i

(

Accendices l A - Summary of 5urveying Procedures for Heave and 5ettlement Measurement at Plant Vogtle B - Heave Analysis i

1 l

Page 3

yi t

i l

i PLANT VOGTLE SETTLEMENT REVIEW i t

i

1. INTRODUCTION AND

SUMMARY

This report presents the results of a review of settlement of major structures at plant Vogtle. The primary purpose of the review was to resolve an apparent discrepancy between estimated and measured settlement.

The rev ew resulted in the fellowing general conclusions:

i

a. Surveying equipment and procedures used to measure excavation heave and building settlement have been in accordance with accepted practice, and the recorded results are reliable within the cegree of accuracy normally expected for such surveys.

b. Previous settlement estimates considered compression of the dense sand underlying the marl bearing stratum to be negligible. Revised settlement analyses presented in tnis report include contributions from both the lower sand and the marl bearing stratum. The revised settlement estimate compares well with measured settlement.

c. Measured heave data, when corrected for depth and loading effects, compares well with revised estimated and measured settlement.

d. Settlement of structures to date has been uniform, has occurred in a normal manner in response to loading, and is of acceptable magnitude. Total and differential settlement to date plus anticipated additional settlement is expected to be well within acceptable limits without adverse effects on structures or equipment.

Page 4

• 1; I

. .- 1 I

2. ESTIMATED SETILEMENT 2.1 PSAR Estimated dettlements Based on the Iield and laboratory test data developec for the Plant Vogtle site during the PSAR investigations, a settlement estimate was made Ior the power block structures for inclusion in the PSAR. The results of the settlement analyses performed in 1973 for Units 1 & 2 are shown on Figure 2.5-16 of the PSAR (8).

The predicted settlements shown on that figure are net -

settlements ontained by subtracting the estimated rebound of the marl stratum from the calculated total settlements. The PSAR predicted total settlements 1 excluding recouna) for Units 1 & 1 are summarized in Tabie 2-1.

The predicted total settlement for the turoine building was significantly lower than for the other structures.

This is because the turbine building is supported on a significant thickness of sand backfill, and therefore, a substantial amount of recompression of the marl was assumed to have occurred prior to start of construction of the turbine building. The control building, Page 5

.lf containment buildings, and fuel building were assumed to be supported on the marl or on lean concrete fill down to the marl stratum. The weight of the lean concrete fill, which was subsequently replaced by compacted backfill in the final design, was included in the total loads used for these buildings in performing the settlement computations. The NSCW towers and the auxiliary building were assumed to be and are supported directly on the marl stratum.

The sand stratum underlying tne marl stratum at the Plant Vogtle site extends to bedrock and is estimatea to  ;

be approximately 1,00C feet thick. Standard Fenetraticn 4

Test criving resistances in this stratum varied from 70 to 100 blows per foot for the depth investigated by soil borings. The majority of the driving resistances indicated refusal of the split spoon sampler: 1.e., the l blow counts exceeded 100 blows per foot. Seismic l

l investigations to a depth of 140 feet into the lower sand stratum yielded compressive wave velocities for the sand ranging from 6400 to 6800 fps. The measured shear wave velocities for the sand ranged from 1600 to 1800 7

fps. Thus, the blow count and seismic data both 1

indicated that the lower sand stratum is very dense.

Therefore, in estimating the PSAR settlement, it was i

Page 6

,Y l

I 1

assumed that most of the settlement would occur due to ,

compression of the marl stratum and that the ,

contribution of the lower sand scratum to the total settlement would be negligible.

The PSAR Settlement estimate considered both elastic settlement and consolidation settlement in the marl stratum. Elastic settlement is defined as the vertical component of movement that occurs when the marl stratum is subjected to a load and no change in voiume occurs as a result. The elastic settlement was estimated using elastic theory and by assigning appropriate slastic i parameters to the marl stratum. The consolidation settlement is associated with volume change occurring over a perloc of time resulting from squeezing cut of the pore water f rom tne marl stratum and dissipation or excess pore pressure caused by the applied load.

Consolidation settlement was estimated using the tr.ecry of consolidation, and was, based on the rebounc portion of consolidation curves from tests on undisturbed marl samples. The total estimated settlement was cbtained by taking the sum of elastic and consolidation settlement in the marl stratum.

T Fage 7

e l-Details of the settlement analyses are given in Appendix 2C of the PSAR, including procedures used for the settlement analysis and the basis for selecting the properties of the marl for use in the settlement '

analysis.

The settlement estimate given in the FSAR was revised to account for some changes in the power block structures dimensions and loadings that were made after the preparation of the PSAR. The revision also considered heave data that had been obtained during excavation of the ocwer block area.

The results of the revised ,

I settlement analyses were included in tne FSAR.

2.2 F5AR Estimated Settlements The range of predicted settlements tor the power clock structures at Plant Vogtle calculated in 1977 is shown en Figure 2.5.4-8 of the FSAR, and the settlement analysis is described in section 2.5.4.10.2 t4). The estimated settlements shown in the FSAR are summarized in Table 2-1.

The range of settlements given in the FSAR is based on upper and lower bound values of Young s modulus for the Page d

t:I l

1 i

i marl stratum. The lower bound value of 4000 KSF was obtained from the PSAR and is based on an empirical relationship between modulus and undrained shear strength (E = 400 x undrained shear strength of 10 KSF).

The value of 10,000 KSF was an upper limit derived from the heave data and insitu pressuremeter and seismic velocity data for the marl stratum. Both elastic and consolidation settlements were considered in estimating total settlement as was done for the FSAR estimate.

The FSAR settlements were estimated using tne SEP0L computer program developed at the Massachusetts j Institute of Technology. The computer program determines stresses and strains induced by surface loads at designated depth intervals in a given soil profile.

For the Plant Vogtle power block settlements, the sanc backfill and the marl stratum were modeled as a layered soil system. Strains were determined at the midpoint of each layer. Elastic settlement was calculatea by l multiplying the calculated strain in the layer by the layer thickness. Consolidation settlement was obtained j using the calculated vertical stress and the rebound 1

j portion or the appropriate consolidation test curve.

The settlement of each layer was determined by taking the sum of elastic and consolidation settlement for the l

l l

Page 9 l

.I

. . i i

layer, and total settlement was calculated as the sum or settlement contributed by each layer.

For structures supported on backfill, settlements were corouted based on compression of the backfill and the -

marl stratum. The auxiliary building and the NSCW towers are supported on the marl stratum; therefore, for these structures it was assumed that only the marl stratum would contribute to the settlement. The contribution of the lower sand stratum to settlement of the power block structures was considered negligibic as in the FSAR settlement estimate. .

3. REVIEW 0F MI.\SUF.ED SETTLEMINT 5ettlement of structures has been measured and recorded since 1979 following pro]ect procedures s1). Over 275 settlement observation markers have been established at locations indicated on project drawings (2). The most recent records available for review include settlement measurements through June, 1984 (3).

A summary and evaluation of leveling procedures for heave and settlement measurement at the Vogtle Plant site is presented in Appendix A. The evaluation was Page 10

{

performed in order to determine the reliability of i settlement measurements before using the measurements to compare them against predicted settlement or for assessing the effects on structures. Based on this evaluation, the elevation of settlement points determined by any individual reading from the time of the initial settlement measurements in 1979 to about August, 1982 can be expe'cted to be accurate within 11/4 inch. Therefore, total settlement of any point relative to an initial elevation determined during this period will have a nominal accuracy of 2 1 /4 inch even if subsequent surveys were performed with higher accuracy. i The total settlement of any settlement point whose initial elevation was determined during the period August, 1982 to June, 1984 (date of most recent available data) can be expected to be accurate within about 11/8 inch. This accuracy also applies to the difference in elevation of any single point determined by two readings during this period.

A summary of total and differential settlement of ma]or structures is presented in Table 3-1. The period of observation for each structure generally covers the time from completion of the basemat to most recent available survey records. The ultimate gross loading and Page 11

I' i

t i

E estimated percent completion at the enc of the period of observation is also noted. Table 3-1 includes the maximum, minimum, differential, and average settlement as well as maximum tilt for each structure. Contour plots for those structures with measurable differential '

settlement sturbine, control, fuel, and auxiliary buildings) are presented in Figures 3-1 through 3-3.

Review of this data indicates that all structures have experienced relatively uniform settlement with differential settlement consisting of a minor tilt.

Where differential settlement has occurrec across a structure, the settlement is less, as expected, in the -

vicinity of unit 2 where construction is still underway.

A summary or differential settlement between structures at a number of representative pcints is presentec in Table 3-2. These points indicate that only minor differential movement has occurred between structures.

4. REVISED SET. TLEMENT ESTIMATE A comparison of measured and estimated (FSAR) settlement for the major structures can be made by examination of Table 2-l. This data indicates that several structures have exceeded or are close to exceeding the maximum Page 12

.I predicted total settlement values stated in the FSAR.

These structures, all of which are founded on or close to the marl bearing stratum, include the auxiliary building, Unit 1 and 2 containment buildings, and the NSCW towers. Since the measured settlement could not be accounted for by compression of only the mari stratum, it was concluded that the lower sand stratum must also be contributing to the settlement. In order to resolve the apparent discrepancy between estimated and measured settlement, a revised settlement analysis including the effect of tne lower sand stratum was perrormed.

As in the FSAR and FSAR estimate, both elastic and consolidation settlements of the marl stratum were taxen into consideration in the revisec estimate. The primary difference between the FSAR settlement estimate and the revised settlement estimate for the Plant Vogtle power block is incorporation of the effect of compressibility of the lower sand stratum. The results of the revised settlement estimate are summarized in Table 2-1.

Elastic settlements were calculated using elastic theory and the following soil parameters for sand backfill, the marl stratum, and the lower sand stratum:

Page 13

,1F

~

i Soil Farameter Backfill Marl . Lower Sand Stratum  !

Moist Unit Weight 120 - -

(PCF; Saturated Unit 130 115 115 Height (PCF) ,

Submerged Unit 68 53 53 Weight (PCF)

Poissons Ratio 0.4 0.5 0.4 Young's Modulus 1500 10,000 See Figure 4-1 (KSF)

- The basis for selection of the parameters for the backfill and marl is explained in Paragraph 2.5.4.10.2 of the FSAR. Values of unit weicht and Poisson's ratio I

for the lower sand stratum are obtain d f rom Table 2.5.4-4 of the FSAR. The values of Young s modulus for the lower sand stratum (Figure 4-1s are based on shear wave velocity data and empirical correlations between shear wave slow strain) Young s modulus and higher strain static Young s moculus. Using tne available j shear wave velocity data for the upper 140 feet of the lower sand stratum (Table 2.5.4-5 FSAR), a plot of low strain Young's modulus versus depth is obtained as shown on Figure 4-1. Values of the low strain Young's modulus i

l below a depth of 140 feet in the lower sand stratum were obtained by extrapolation using the Seed-Idriss relationship (Reference 9):

. ^f I

Page 14 L

I I

i i

Gd = 1000K ' "# !

2 where K is a parameter that depends on the void ratio 2

and the strain amplitude of motions.

0$isthemeaneffectivestressandG id is the low strain shear modulus in psf. The low strain Young s modulus iEd ) is obtained from the equation E

• 2'1*"'O d d

Where v is the Poisson s ratio of the icwer sand stratum.

Based on studies made by Swiger tReference 10>. the static Young s modulus for the lower sand stratum may be obtained by taking it equal to about 1/3 the low strain value (E d ). Therefore from Figure 4-1, the static Young's moduli of the various layers in the lower sand stratum are as follows:

Page 15

[

deoth Below Grade iFT) E iK5F) 160-260 10,800 ,

260-460 13,500 460-760 17,500 760-1160 22,000 The lower sand stratum is approximately 1000 feet thick based on geophysical data obtained during the Millett Fault Study in 1982 (Reference 11). For the settlement analyses, the lower sand stratum was divided into four layers of thickness, 100, 200, 300 and 400 feet, respectively. Each layer was assigned an appropriate j value of Young s modulus as shown above. Total elastic settlements were obtained by adding the elastic settlements obtained in the backfill, marl, and the various layers of the lower sand stratum. Elastic settlement of each layer was calculated by multiplying the strain t'zi in .ne layer by the height cf the layer.

The strain t'zi is given by the expression

'* =

{ *z - V{ 0 x+Oy 1 whe re 'z , 'x and'y are the principal stresses induced at center of the layer by the loads, E = Young s modulus, Page 16

.10 l

1 I

and y = poisson's ratio. The stresses 0z, 0 x and 8 y were obtained at the midpoint of each layer from the SEp0L f t

computer program. In calculating stresses and I settlements, both the weight of the backfill and structure loadings were taken into consideration. --

The consolidation settlement was computed from the equation:

I

1. c - C ch tn (o,#3y )

1 + e, o$

where 9e = Consolidation Settlement ,

t C = Compression index c

h = Marl thickness o = Initial void ratio e

'v = In situ effective vertical stress at the mid-depth of the marl stratum 08v = Additional vertical stress at mid-depth of the marl stratum due to the surface load.

The thickness of the marl stratum was taken as 70 feet for purposes of computing consolidation settlement.

Considering the highly preconsolidated nature of the marl stratum, the compression index (Cc ) used in the above formula was taken equal to the rebound index LCr' obtained from laboratory consolidation tests.

Page 17

IT i:

i-i:

Total settlements shown on Figure 4-2 include both elastic and consolidation settlements of the marl >

stratum and elastic settlements of sand backfill and the lower sand stratum. Since the mari stratum is highly preconsolidated, consolidation settlements are very small compared to elastic settlements.

A comparison of measured and revised estimated I

settlement for the ma.1or structures can be made by examination of Table 2-1. These cata indicate that settlement of all mayor structures, as of the date of j the latest measurement, is well within the revised estimated value. Some additicnal settlement is anticipated due to additional structure loads and weight of fill yet to be placed. However, measured settlement to date plus this additional future settlement is expected to be less than the revised estimated settlement for all structures. This is because the revised settlement estimate is based on conservative values of the consolidation and elastic properties of the supporting soils, page 18

e

,L i

5. CONCLUSIONS The review of settlement of structures at Plant Vogtle resulted in the following conclusions:

a. Measured settlement is reliable within the degree of accuracy normally expected for such surveys, i.e., 1 1 /8 inch to 1 1 /4 inch (see Appendix As.

b. Measured heave data, when corrected for depth anc loading effects, compares well with measured settlement tsee Appendix B).  ;

c. The results of the revised settlement estimate, which accounts for compressibility or the 1:wer sand as well as the compressibility of the marl, compare well with maximum recorded settlement plus anticipated future.setriement.

d. The primary factor contributing to settlement at plant Vogtle is the weight of fill over a large area causing compression of the underlying marl and sand strata. Therefore, while total settlement recorded to date is larger in magnitude than originally estimated, the settlement of fill and 1

Page 19

1[

i structures is quite uniform. Differential f settlement therefore, is expected to be small and comparable to values originally estimated. Since it is primarily dt.fferential settlement that can affect structures and connections between structures and since differential settlement is

, expected to be small, no adverse effects on structures or equipment are anticipated as a result of settlement at Plant Vogtle.

6. REFERENCES  ;-

1. Vogtle Project 5pecification No. X2AF01, Section No. C10.1, Obtaining and Recording Foundation Settlement Data.

2. Vogtle Project Drawing No. AX2D55V001, Settlement Observation Markers Location and Detail.

3. Vogtle Project Settlement History, Letter C4380, dated August 9, 1984, File X2BA08.

4. VEGP, Unit 1 and 2, Final Safety Analysis Report.

'I Page 20

s'

5. Goldberg, Zoino, Dunnicliff and Associates letter to BPC, 8/22/77 (Re: Instrumentation ror Heave and Settlement Measurementss.

6. "Use of Extensometers at Palo Verde Nuclear Station," B. M. Ghadiali and W. B. Tijmann, Journal of Construction Engineering, ASCE, C01, March 1981.

Pages 49-59.

7. Grand Gulf Nuclear Station, Units 1 and 2, FSAR (Section 2.5.4.5.4 and Figures 2.5-76 through 2.5-80).

8. VEGP, Unit 1 and 2, Preliminary Safety Analysis Report.

9. Seed, H. B. and I. M. Idriss " Soil Moduli anc Damping Factors for Dynamic Response Analyses" EERC, University of California, Berkeley, CA, December 1970.

10. Swiger, W. F. " Evaluation of Soil Moduli for Soil Structure Interaction Analysis" Presented at the Conference on Analysis and Design in Geotechnical Engineering, Austin, Texas, 1974.

Page 21

i i

11. Studies of Postulated Millett Fault, Report Prepared for Georgia Power Company, by Bechtel Power Corporation, October 1982.

Page 22

i I

i TABLE 2-1 EST!!mTED #6 RECORDED SETTLEMNT PSAR FSAR Revised MaximunRecorded Estimated Estimated Estimated Settlement Structure (in.) (in.) (in.) (in.)

Turbine Building Center 0.8 0.7-1.6 2.5 1.6 Corners 0.8 0.9-1.8 2.3 1.7 -

Auxiliary Building Center 1.6 1.0-2.7 4.4 3.0 Corners 2.5 1.4-2.7 4.6 3.2 ContainmentBuilding(Unit 1)

Center 3.0 1.5-2.5 4.3 3.0 Edges 2.8 1.4-2.2 4.0-4.2 3.1 Containment Building (Unit 2)

Center 3.0 1.5-2.5 4.3 2.5 Edges 2.8 1.4-2.2 4.0-4.2 2.7 Contrel Buildin3 Center 2.5 1.0-1.9 3.3 1.5 i Corners 2.6 1.2-2.1 3.2 2.1 Fuel Handling Building Center 2.8 1.3-2.4 3.4 1.9 Corners 2.7 1.3-2.7 3.4 1.8 NSO4S Tower 1A Center 2.2 1.3-2.6 4.7 3.4 Edges 2.5 1.4-2.6 4.5 3.4 l

NSO4S Tower 18 Center 2.2 1.3-2.6 4.8 2.8 Edges 2.5 1.4-2.6 4.5-4.7 2.8 NSO4S Tower 2A Conter 2.2 1.3-2.6 4.8 WM Edges 2.5 1.4-2.6 4.5-4.8 WM NSO4S Tower 2B Center 2.2 1.3-2.6 4.7 3.0 Edges 2.5 1.4-2.6 4.5-4.8 3.0 Diesel Generator Building Center WA 0.5-1,3 2.0 WM Corners WA 0.3-1.2 1.1 WM

• Not Available

D i

p TABLE 3-1 S* 0WtY J OF RECORDED SETTLDENT Period of Observation Estimated Ultimate Percent RangeofRecordedSettlement(1) Max. Tilt Structure Free Te GrossLeading(KSF) Ceaplete Max. Min. Diff. Avg. (Radians)

Unit 1 Turbine 81dg. 3/7/90 5/29/M 3.2 100 1.8 1.3 0.4 1.5 0.00006 Unit 2 Turbine 81dg. 3/7/90 5/2S/M 3.2 100 1.5 0.9 0.6 1.2 0.00013 ,

Auxiliary Bldg. 7/5/79 Y17/M S.S 100 3.2 2.5 0.7 2.8 0.00018 Contrel Bldg. 3/4/81 F12/M 4.7 100 2.1 0.9 1.1 1.5 0.00032 Fuel Bldg. FF81 6/15/M 6.3 100 1.9 1.5 0.3 1.7 0.00012 Unit 1 Containment 1/19/80 3/21/M 8.4 95 3.1 2.9 0.2 3.0 0.00020 Unit 2 Containeent 1/19/80 V21/M 8.4 85 2.7 2.3 0.4 2.5 0.00033 NS0 6 Tower 1A 3/24/80 5/15/M 10.3 100* 3.4 3.4 0 3.4 Nil NS06 Tower 18 1/10/80 5/15/84 10.3 100* 2.8 2.8 0 2.8 Nil

~

NS06 Tower 2A 10.3 70*

NS06 Tower 2B 3/24/80 5/15/94 10.3 70* 3.0 3.0 0 3.0 Nil

• Excluding weight of water (1)Settlementininches l

,T

• s TABLE 3-2 DIFFEADRIAL SETTLDGE BEMEN STRUCTIRES Differential Settlement Marker Period of Observation At End of Period of No. Free Te Observatten(in.)

Structure S/13/81 W84 6.15 Centairment Unit i 125 Fuel.andlingBldg. 151 S/13/81 6/5/M 0.15 Fuel Handling Bldg. 151 Aux. Bldg. 120 4/2S/83 2/23/M 0.15 Aux. Bldg. 187 NSO4S Tunnel 1B 188 6/5/81 3/21/M 0.20 Contrel Bldg. 122 Containment Unit i 123 10/S/81 6/21/ 83 0.20 Aux. Bldg. 233 Fuel Handling Bldg. 257 2/1/M 0 226 3/15/83 Containeent Unit 2 j Control Bldg. 222 6/5/84 0 Contrel Bldg. 223 10/3/81 Fuel Handling Bldg. 229 10/9/81 3/7/84 0.38 Containment Unit 2 426 Fuel Handling Bldg. 258 3/7/84 0.10 426 10/9/81 Contairment Unit 2 Fuel Handling Bldg. 230

.T I

ed IN  :

W M14

1. h G

.ad w

M O

O O M G

C/3 e 0 1 9

e4 e o P E

e e g

Z N

m J 9 0 H

H i

W

'A e e  !

1 E

ad lN .Pe D Z C D g . - _ _ _ y.1 E

D H 4 "l4 G3

-e e N l N M  %

A

= l M

3  : e e

'O O I* \ #

N N

$ # M

\. .

G N

H m

E C

,O w

US G 9 .42 Q

W l 0

. e

• TE

.O4 bd

1. . 9 z \

i

c' r4 i

. S.

M:4 -

G M N Jd "M IN be e=e f

M C

y W w M

k.

M G

t/3

.-s I4 M

,I M

N Z

D

!E a

23

.J .

32 "

f/2 (A H

.=

C Z Z 3 0

d -

11 V

4

= .,

2 C

3 M-C N M

H Z

C O -

M N 4 M

8

/  %

C O O

• w I W M I ("J I

N >

6.

. b m u

a E o

, O l

t 9 O

w be 0

n.

e I

I i

/

t Z

u r

e k

4 r 4 8 a 8 / M

/ 7 9 1 t 1 / n

/

5 e 6 m

.3 - e ,

- l 9 t 1 7 t 8 / e

/ 5 S 5 /

/ 7 -

6 N

• 4*

G N

I .

D 1 L

I T U . i I B . N

• U Y

• i R Y A j5 R I A L

I l I

L X * = I ll X AT U N A S DE T f

NM I N

AE 3z L I U

LT .3; ET 2 iE l

s I o I . ' ,

FS

/e/

i 3

3 L

E

/ I s

'3 i U

F

/

G I

F / .

L* [ "

2 T

I 1 2 2 N

/ U i

/

N

.J

>s i

1 l

t I

l E (KSF)  !

20x103 40x103 60x103 80x10 3 i t t t i SAND BACKFILL MARL 200 - LOWER SAND LAYER 1

\

LOWER SAND LAYER 2 400 - 1 1

6 i

O h.

600 - LOWER SAND LAYER 3 s

5 u

3

• 800 -

=

$ 1

$ Es = 3 ED LOWER SAND LAYER 4 I

t 1000 - i Ep I

u 1200 -

1400 -

FIGURE 4 -1

L.

D I

S E S

3

L,2 D . E; L

1

.4

. C 0 C

E M

{

1 ,

2 3

4e 5

e .4 7

3 2

,2 3 4 3 4 2 4 4. 6

, 0 3U 8 6 C O

N N

' T v 4

1 I

T .* 3 A

I N

.4 1

2 M E

F N i

4 T g

u 4 2 r o .4 -

e 5 8 3

4 .

- 4 2 A 3 T N .1i R 4 S U 1

d C O

U R

C N B e -

W 1 T I v 8 L i 4 3 3 R 3 2 N 2 n T -

I A e .4 -

f. *

' O . 0. e .2 E e O 4 R

4 3 L 3 5 5 4

. d W Y

. B S

E . B L e R 3 L D S B

_ t L -1i 2 D C t

l D 3 C

_ e C .4

. m e 4 n

5 e .4 8 4 t

2 O

C[ .4

_ U N

_ N T 4 A

,4 7

1 I

T .' 3 I

N 1 M

g1 E

N

5. T

. 4 6 4 0 3 6

2

,3 2 4 , .4 2 .,3 2

_ 5 7 D 3 I

_ E E 2 S 1

EO M,0 4 L

_ e 4 4

5 C E

2 y N T

o t

a l

s e

t t

l e

m e

n

_ t

_ i n

i n

c h

_ e n

,lf APPENDIX A

SUMMARY

OF SURVEYING PROCEDURES FOR HEAVE AND SETILEMENT MEASUREhENT AT PLANT V0GTLE l

l l

l l

l I

j so l

SUMMARY

0F SURVEYING FROCEDURES FOR HEAVE AND SETTLEMENT MEASUREMENT AT PLANT V0GTLE All surveying for the purpose of heave and settlement measurement at the Vogtle plant site since the beginning of construction in 1974 has been performed by Georgia Power Company. Leveling at the site can be separated into 4 phases based on nominal accuracy achieved. The accuracy achieved is based on leveling equipment anc procedures useo during the time period. The first phase covers the period 1974 to 1977 when heave measurements were made during site excavation. The second phase covers initial building settlement readings beginning in 1979 and continuing to mid-1962. The third phase covers recent settlement readings from mid-1982 to the present time. The final phase will apply to settlement measurements beginning in mid-1984. The quality of surveying equipment and procedures has increased in each subsequent phase as summarized in Table A-1 and described below. Surveying to the tops of the heave points during 1974-1977 is believed to have been accomplished with a nominal accuracy of +1/4 inch at best; in contrast, surveying to building settlement points beginning in mid-1984 is expected to be accomplished with an accuracy of

+1/16 inch.

A-1

f

~

.a Heave measurement (6/74 - 8/77)

Levels were run to the heave points over a circuit approximately 1.7 miles in length beginning and ending on the deep benchmark (GZD-1). A Zeiss NI2 t2nd orders level without micrometer was used. A yard rod s3rd order at bests was used with direct readings to v.01 feet with the nearest U.001 feet estimated. Turning points consisted of wood stakes. .

The observing procecure used was 3-wire leveling where the rod reading it obtained by taking the mean of three horizontal cross-hair readings thorizontal and two stadia cross-hairss. No attention was paid to balancing backsights and foresights, and some sights were unbalanced to the extent of 25 feet vs 200 teet. No corrections ror curvature, refraction, temperature, or other factors were applied. A ist order closure tolerance was followed t0.017/$), but no adjustment for misclosure was made. Heave point elevations reported in the Goldberg, Zoino, and Dunnicliff report of August 22, 1977 are based only on the levels run out from GZD-1 to the heave points. Adjustment A-2 i

L

r 1

for misclosure back to GZD-1 was recently made by GPC based on the original field notes. The misclosure adjustment was based on the number of turns (setups), and adjusted elevations for the 9 heave points reported (telecopy from C.

Chiappetto dated May 25, 1984). -

The results of the heave point leveling surveys during this period are considered to be 3rd order or less, with a nominal accuracy of about 1/4 inch to 13/8 inch.

i m

, Early settlement readings (7/79 - 8/82)

Levels run to building settlement points during this period were based on control monument CM-1. The distance from the l control monuments to the buildings was double run, and misclosure limited to 0.017/E. CM-1 was checked against l benchmark GZD-1 approximately every 90 days. Settlement points were included in circuits beginning and ending on the control monuments. Loop misclosure was limited to 0.017/M and any misclosure was adjusted over the entire circuit according to the number of turns (setups). Leveling from plant grade outside buildings was done by optical leveling either with a leveling rod for small differences in A-3

o I elevation or by direct reading on a tensioned invar steel tape held vertically as plant grade was raised. No temperature or other corrections were made for the vertical taping. Settlement points generally consisted of brass caps embedded in slabs or, in some cases, short pieces of rebar -

grouted into walls. Metal rods were used for turning points.

The Zeiss NI2 level continued to be used, but with a micrometer attachment. Two, double-scale 2nd order rods were used which, in combination with the micrometer, allowed direct readings to 0.001 feet with the nearest 0.0001 ,

estimated. The 3-wire observing procedure was used, and backsights and foresights were roughly balanced. No corrections for curvature, refraction, temperature, or other factors were applied. Initial elevations for any settlement points established during this period were determined by averaging 3 independent leveling runs, each of which met the misclosure tolerance.

Surveying equipment and procedures improved over the i

previous period. However, potential errors due to vertical taping without temperature correction, use of rebar settlement points, adjustment over the entire circuit without regard to the relative accuracy of different A-4

I I

i sections, and neglecting ccrrecticns limits the cverall surveying accuracy to 3rd order. A nominal accuracy of :l/4 inch was probably achieved during this period.

F.ecen t settlement readinas (8/82 - 9/84)

Levels run to the building settlement poin:s have been c:ut'e-run in a manner similar to that dmring the previcus peried using CM-1 and CM-2 as benchmarks. Mcwe /er , :he 3-wire .eveling prccedure was abandcned based :n :he -

t rec:mn.endaticn c: GFC's surveying censui: ant . Earl Cudley, X.'i ~nstrumen:s represen:ative, Eirmingham, AL. ::he advised

hem : hat instrument :stics for the stadia hairs are nct as ac:ura:e as Icr the hori:cntal cr:ss-nair. The :-wire precedure was replaced by the folicwing cbserving peccedure at each setup:

Backsight - read horizontal cross-hair and stadia intervals (mean of stadia readings used as check against misreading of horizontal cross-hair).

A-5

,1 t

i l,

}

Fcresi?ht - read horizontal cross-hair and a.adia ,

i i

intervals (mean of stadia readings used i only as check against misreading of horizontal cross-hair)

The Zeiss NI2 level with micrometer attachment continued in use. Seccnd order rods were replaced with 3 meter, 1st order :alibrated fods with spring-mounted invar s: eel tape f:r runs between the control monuments and the ben:hcark..

Seccnd ceder rsds c:ntinued to be used f:r the settlement 't scin.: eleva:i:n measurements. Sight distances and c:her tolerances have been maintained within 1st crder requi.emen:s as prescribed in :he NOAA Manual NOS NG2 3, Geodeti: Leveling t1331). Other imgrsvements in:luiei the use of ;i!.ld steel turning plates, Nelscn stud se:tlement points in walls, temperature correcti:n ::r vertical tapin?,

and the application of correcticas for curvature, refraction, rod index error, and collimation.

3 6

A-6

1[

).

Initial elevations of settlement scints established during this peri:d were determined by averaging 3 runs as during the previous period. The observing sequence was revised to include the building settlement points as spurs rather than -

including them in the level circuit. Figure A-1 indicates schematically the procedure used for running a typical level circuit from a permanent benchmark (control monument) to settlement monitoring points within a structure. points 01 thrcugh 06 represent settlement monitoring points censisting primarily of brass caps embedded in the floor or, in some cases, toits projecting fr:m walls. Leveling between CM and ..

IEM', and between IBM 2 and the settlement menitoriny points, is acccmplished by conventicnal techniques with a spirit leve'. and leveling rod. The secticn between TEMI and ThM:

invc;ves leveling from a temporary benchmark at the plant grade elevation outside the building to another temporary benchmark located in the basement of the building. The difference in elevation between TBM1 and TEM 2 may be as much as 100 feet, and is measured by direct reading of elevations l

on a calibrated steel tape suspended vertically with a 10 l pound weight attached at the bottom. The tape is calibrated for both the condition of the weight and for temperature, and appropriate corrections are made to the measured data.

The sequence of observing is as follows:

A-7

^

llf a

l I

CM to TBM1 to IBM 2 to vi to 02 to v3 to 04 to 03 to 02 i to 05 to 02 to 06 to TBM2 to TEM 1 to CM  :

Recognizing that the least accurate operations are taping the large vertical distance and leveling to settlement points in congested areas in the buildings, adjustment for misclosure is made as follows:

The mean of two 1-way runnings f om CM to TBM1 and TBM1 to CM is determined, and the elevation of TBM1 thus escablished. Similarly, the mean of two runnings between TBM1 and TBM2 is used to establish the elevation [

of TEM 2. TBM2 to 01 to 02 to 06 to TBMI is considered a loop, and any misclosure is proportioned according to distance around the loop to establish the elevations of 01, 02, and 06 from TBM2. Points 03, 04, and 06 are considered spurs and any misclosure is adjusted between them.

Leveling during this period is considered to be 2nd order with a nominal accuracy of 21/8 inch.

A-8

ik Future settlement reading i beeinr ing 8/84)

Beginning in August 1984, all building settlement surveys will be ac:omplished using ist order equipment and procedures. The Zeiss NI2 level will be replaced with a Wild N3 1st order level with built-in micrometer. This instrument has already been used for a 39-mile, 1st order loop in late 1983 that tied the plant site control monuments to the National Geodetic Survey tNG5) ist order network at Waynesboro. The results of the GFC survey or the Waynesboro line have been submitted to the NGS for inclusion in the NGS 1st order network. The Wild N3 level is also presently

• being used to monitor settlement points on the natural draft cooling towers.

Using the Wild N3 level and 1st order rods, readings are made to 0.0001 feet and estimated to 0.00005 feet. Sight distances, closure, and other tolerances will be in accordance with 1st order, class II requirements of the NGS 3 manual. All corrections necessary for ist order leveling prescribed in the NGS 3 manual will be applied. The observing sequence and procedure will be the same as during the previous period with the exception that adjustment for misclosure error will be by distance rather than number of setups. Two, 3-meter long, 1st order rods have been A-9

s[

e modified to a 2-meter length for use in buildings which should improve the accuracy of leveling in congested, low headroom areas.

Leveling using the above equipment and procedures should meet the requirements for ist order, class II leveling as defined by the National Geodetic Survey. The nominal accuracy should be approximately 2 1/16 inch.

Conclusions The elevations or heave points measured during the period 1974-1977 can be expected to have an accuracy of gli4 inch, at best, due to survey error alone. Further inaccuracy could be expected within the heave instrumentation itself (cod coupling error, obstructions, temperature effects, misreadings, etc.).

The elevation of settlement points determined by any individual reading during the period 1979 to August 1982 can be expected to be accurate within 11/4 inch. Theretore, total settlement cf any point relative to an initial elevation determined during this period will have a nominal accuracy of 11/4 inch even if subsequent surveys were performed with higher accuracy.

A-10

i!

The total settlement of any settlement point whose initial elevation was determined during the period August, 1982 to August, 1984 can be expected to be accurate within about 11 /8 inch. This accuracy also applies to the difference in elevation of any single point determined by two readings during this period.

Elevations of settlement points measured during August, 1984, and later are expected to be accurate within 1 1 /16 inch. However, the accuracy of total settlement relative to the initial reading will depend on the date of the initial  :

reading as discussed above.

1 a

A-11

TABLE A-1

SUMMARY

OF SURVEYING PROCEDURES Heave measurement Early settlement readings Recent settlement readings Future eettlement readings Period:

(6/74 to s/77) (7/19 to n/n2) (s/s2 to s/84) (after s/s4)

Equipment Zetse N12 (2nd order level). Zeiss N12 level utth Wild N3 level (let order) no micrometer mictameter attachment r 2nd order reus let order rod (double rod, Yard rod (3rd order or less)

(let order rod BM to CM's) douhle scale) read .Ol'. est. .001' read .Emil'. est. .(M Wil '

ead .0001' est. .00005'

~

Monuments Eft =CZD-l BM=Qt-I and CM-2 t (2 mile circult) (1/2 mile one w y) wood stake turning pts. metal rod Tr's Wild turning plate Nelson stud sett!. points  ;

reber settlement points 7

brass cap floor points observing 3-wire leveling double scale,1/2 stadia r Procedure sight diet. not balanced eights roughly balanced sight diet. per NGS-3 sight diet.. closure per NCS-3 manual mensal for ist order, T class !!

let order closure (.017 M) r level circuit double run Processtag misclosure adjustment stoclosure adjustment (number of setups) 2 (by distance) no corrections appiled 1 correction for curvature, refraction, a!! corrections per NGS-3 rod indes, colltestion manual no temp. correction (vert. taping) vert. tapte.c corrected f or temp. r

~

Accuracy 3rd order or less -3rd order 2nd order Ist order, class 11 moninal *

(et/s-) (el/86-)

accuracy .

a sisigle (el/4 to .3/a-) ' (el/4-)

point ,

t . .-

q

,I' YFP/ CAL GETTLEMENT su2VEY lE1/EL C/RCUlT \

FIGU2E A-l CM U o/ sz 07 o o_ -o .

I npyg

+

-+ Q h f O O O og of 04 PL AN CM 7BMI

.T

~

\,

Q'w I

, \ _,

7BM2 a p.. .

mh 1

. A. A.

.El E VA Tl0W

,1f i

i h

t 9

APPENDIX B HEAVE ANALYSIS i

a HEAVE ANALYSIS Nine heave points were installed and monitored during the excavation phase from June, 1974 to August, 1977 (Reference 5s. The locations of the heave points are shown in Figure B-1. Heave of the marl bearing stratum, measured during excavation, ranged from 0.6 to 1.7 inches with an average of 1.3 inches treference 4, Section 2.5.4.10.4 and Table 2.5.4-13s. The measured heave data has been corrected for minor survey adjustments and re-analyzed to correlate j measured heave with excavation history, heave point installed elevation, and approximate original ground elevation. The resulting heave data is summarized in Table B-1. Heave points 4 and 6 appear to be unreliable and are not considered in any further analyses tthese points were also discounted when reporting heave in the FSAR, Reference 4). Measured heave for the remaining 7 points judged to be reliable ranges from 1.0 to 2.3 inches with an average of 1.4 inches. This compares well with the average of 1.3 inches of heave previously reported in the FSAR (Reference 4).

B-1

,W o

Measured heave, however, cannot be compared to settlement upon recompression until several corrections are made. The first correction is for depth of heave point below the base of the excavation since the heave at plant Vogtle was measured at depths of 9 to 27 feet below the base of the -

excavation stop of competent marls. The second corr,ection required is for depth of overburden removed at the time of installation of the heave point since measurement of some heave points did not commence until excavation was well under way. The third correction required is to account for the increased density of backfill compared to the density of the natural material removed. ,

f The maximum heave of the base af large excavations occurs near the center of the excavation away from the restraining influence of the sides of the excavation. Furthermore, heave is a maximum at the surface of the base or the excavation and decreases with depth. Measured heave at the Palo Verde Nuclear Generatir.g Station tReterence 6 s is summarized in Figure B-2. Measured heave at the Grand Gulf Nuclear Plant (Reference 7) it summarized in Figure B-3. As can be seen from these plots, haave at a depth ot 100 feet, for example, ranges from 10 to 30 percent of the heave measured at the bottom of the excavation.

B-2

l Since heave was measured only at a single. depth at each heave point locat.'.on at Plant Vogtle, it is not possible to determine a depth correction factor directly. Therefore, data from other sites was used to estimate an approximate value. The average heave vs. depth curves for Grand Gulf and Falo Verde are plotted in Figure B-4. Geological conditions at Grand Gulf are similar to Plant Vogtle thard clay stratum over dense granular material), as are the conditions at Palo Verde (dense sand and silt and hard clay to a depth in excess of 300 feets. The excavations at Grand Gulf and Falo Verde are also comparable in configuration to that at Plant Vogtle. Therefore, the average heave vs. j depth relationship for Grand Gulf and Palo Verce was applied to the Flant Vogtle heave data. The correction factors used and corrected heave data are shown in !able B-2 The heave data, corrected for depth effects, ranges from 1.3 to 2.6 inches with an average of 1.8 inches.

As indicated in Table B-1, the heave measured in the field was not always in response to the full average depth of excavation of 75 feet. Assuming a linear relationship between heave response and depth of overburden removal, the measured heave data (corrected for depth effects) has been extrapolated to an average excavation depth or 75 feet as shown in Figure B-5. The corrected data indicates heave for B-3

,U'

!i 75 feet of excavation ranging from 1.9 to.3.3 inches with an average of 2.6 inches.

Due to the increased density of backfill compared to the density of natural material removed, the weight of the fill -

and structures causing recompression is greater than the weight of overburden removed causing heave. This effect can be approximated by using an equivalent excavation depth of about 90 feet. Again, assuming a linear relationship between heave response and depth of overburden removed, an extrapolation indicating the amount of heave for 90 feet of excavation is shown in Figure B-5. The corrected data ,

indicates heave ranging from 2.3 inches to 3.9 inches with an average of 3.1 inches.

A comparison of recorded settlement or structures founded on the marl and corrected heave (in inches) can now be made as shown belcw:

law _ averace high measured settlement 2.3 2.9 3.4 corrected heave 2.3 3.1 3.9 B-4

IF TABLE 9-1 EASLItED EAVE DATA De NASLItED START Rolle W 8t0LIG W OKRILIt004 TOTAL 1054 W DEFTH KLW W DATE EL.(FT) EL.(FT) DATE EL.(FT) EL.(FT) R9GED (FT) (FT) (1N) FiftlL EXCAtl(FT) RDWtKS 126.24 63 0.09 1.1 24.0 1 F22n4 199 126.15 F 7 n 7 136 124.63 49 0.08 1.0 19.5 2 #1974 193 124.55 2/2& 77 144 126.30 54 0.11 1.3 19.0 (1) 3 F1U74 199 126.19 2/2& 77 145 1M 126.38 63 0.19 2.3 10.0 F1&74 199 126.19 8/7/77 145~ 125.90 55 0.28 3.4 19.0 (2) 4 U22n4 200 125.62 2/2 & 77 125.98 69 0.36 4.3 5.0 U22n4 200 125.62 8/7#7 131 126.22 59 0.13 1.6 19.0 5 U1U74 204 126.09 2/26/77 145 .

125.96 30 0.36 4.3 -

(3) 6 U16n4 199 125.60 7/21n4 169 105.04 71 0.11 1.3 27.0 7 F1974 20 3 104.93 F 5/77 132 123.52 26 0.10 1.2 9.5 8 0Y74 159 123.42 Bn n7 133 124.13 48 0.14 1.7 9.0 9 U30n4 181 123.99 S a n 7 133 Remarks:

(1) Added 2' to 2/2977 for sapet (2) All data suspect since only 1 point in 1974 l

f (3) Not consistent with any other W's i

l l

l t

.T

\

TABLE B-2 HEAVE CORRECTED FOR DEPTH EFFECTS DEPTH BELOW HEASURED CORRECTION HEAVE AT HP EXCAVATION (FT) HEAVE (IN) FACTOR EXCAVATION LEVEL (IN 1 24.0 1.1 0.68 1.6 2 19.5 1.0 0.75 1.3 3 10.0 2.3 0.87 2.6 19.0 1.3 0.75 1.7 5 19.0 1.6 0.75 2.1 7 27.0 1.3 0.65 2.0 8 9.5 1.2 0.88 1.4 9 9.0 1.7 0.88 1.9 k

lf N

H.P. 2 f H.P.1 G Q N 83 + 74.50 i'

.N 83+ 74.50 TURRINE BLOG.

E 99 + 54

• E93+65 H.P. 3 O

N 82 + M.M E N + $5 ,

CONTROL SLOG.

H.P. 4 RE ACTOR CONTAINMENT BLOG. Q RE ACTOR CONT AINMENT BLOG.

N 00 + $4.57 E 96 + 65 H.P. $ H.P. 6 O O N uo + 00 N 80 + 00 E94+95 E 98 + 35 AUXILIARY BLOG. j H.P. 7 O

N 78 + 51.50 E 96 + 85 NUCLE AR SERVICE NUCLEAR SERVICE COOLING TOWER COOLING TOWER H.P 9 H.P. 8 O 1 l l O

N7.. 0 E93+94 V (Mr. 0 E be + 36 i

Ref: FSAR Figure 2.5.4-9 FIG. B-1 HEAVE POINT LOCATION PLAN

O t

t i

i HEAVE (IN.)

0 0.5 1.0 1.5 2.0 0 -

7

/

MPE-2

/

/

/

MPE-3 j

/ ~

/ MPE-1 A

/

DEPTH 100 /

BELOW I BOTTOM OF /

EXCAVATIO:3 .

(FT)

I

~

l I

l (D

?

! t -

'! anchored at depths o

of 300'-346' 200 -

l Excavation width = 400 feet Depth of excavation = 60 feet MPE = Multiple Position Extensometer FIG. 8-2 MEASURED HEAVE AT PALO VERDE 4

I I

4 HEAVE (IN.)

10 1,5 2,0 0,. 5 0

RE-B f E' L-RE-F -

-A

/

j, /

DEPTH .

/

BELOW /

BOTTOM OF EXCAVATION f

/f (FT) /O

/

//

/

/

100- y //

f/

/'

/

' Excavation width = 1000 feet

/

Depth of Excavation = 105 feet anchor 150, FIG. B-3 MEASURED llEAVE AT GRAND CULF

i' l

I i

% HEAVE AT BOTTOM OF EXCAVATION O 50 100 0 -

20 -

40- - /

/

/

DEPTH  ! 5 60 - -

(FT) /

4 80 -

4s 100-FIG. B-4 AVERAGE DEPTH CORRECTION FACTOR J

I I

i i

o l

100-AVG. 90' FILL DEPTH 90- .g 80-AVG. 75' EXCAVATION I

70- D 1 3 n O $- ,0

!;' 60- '

e,-

c 3e r' .

d 50- 2 4 9

i 40-p y 30 ,

f

/ 4

. 8 4

20 @ Af 10, O 0

/ -

i 0 1 2 3 4 5, l

[ HEAVE (IN.)

l I

G= Heave Point FIG. B-5 EXTRAPOLATION OF HEAVE DATA a