ML20072F360

Forwards Responses to NRC Questions Discussed at 830428 Meeting Re Seismic Confirmatory Program.Equipment Margin Analysis for Safe Shutdown Equipment for ACRS Response Spectrum Scheduled After First Refueling Outage

ML20072F360
Person / Time
Site:	Summer
Issue date:	06/22/1983
From:	Dixon O SOUTH CAROLINA ELECTRIC & GAS CO.
To:	Harold Denton Office of Nuclear Reactor Regulation
References
NUDOCS 8306270356
Download: ML20072F360 (123)

Text

{{#Wiki_filter:* m

• s SOUTH CAROLINA ELECTRIC & GAS COMPANY Post orrica 7s4 COLUMBIA. SOUTH CAROUNA 29218 O. W. Onnon, Jn.

Vice PRESIDENT June 22, 1983 NUCLEAR OpenATIONS Mr. Harold R. Denton, Director Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Subject:

Virgil C. Summer Nuclear Station Docket No. 50/395 Operating License No. NPF-12 Seismic Confirmatory Program

Dear Mr. Denton:

On September 1, 1982, representatives of the South Carolina Electric and Gas Company (SCE&G) presented to members of the NRC Staff, a proposed Seismic Confirmatory Program to satisfy License Condition 2.C(25) of the Virgil C. Summer Nuclear Station Operating License. After incorporation of Staff comments on the Program, SCE&G formally submitted its Seismic Confirmatory Program on September 24, 1982. Report I, concerning the experimental phase of the Program, was filed on February 1, 1983. Additionally, an Addendum was filed to the Report on March 9,198 3, and a meeting was held with the Staff on March 10, 1983, to review the Report. NRC questions were transmitted informally and draft responses were presented to the NRC at a meeting on April 28, 1983. In a letter dated May 4,1983, SCE&G received formal questions from the Staff concerning the program. Additional Staff comments and questions were documented to SCE&G in a meeting summary dated June 10, 1983. This letter serves to transmit formally ten copies of the responses to Staff questions discussed on April 28, 1983, and the further questions contained in the meeting summary. The initial questions are numbered 1-6 as in the Staff's letter. The questions labeled 1 and 2 in the NRC's meeting summary are herein labeled 3A and 2A respectively. As discussed in the summary section of our Report, the experi-mental results demonstrate a reduction in the "Monticello enveloping response spectrum." Taking into account this reduction, the equipment margin analysis to be performed for \\ 8306270356 830622 PDR ADOCK 05000395 l P PDR I,10

. g'. Mr. Harold R. Denton June 22, 1983 Page #2 the "Monticello enveloping response spectrum" will be encompassed in the equipment margin analysis to be conducted (M =4.5 spectrum, anchored for the "ACRS response spectrum" L at.22g, as discussed at our ACRS hearings in February and March 1981). Equipment margin analysis for safe shutdown equipment for the "ACRS response spectrum" is scheduled for completion prior to startup after the first refueling outage. It is our understanding, based on conversations with the NRC S*aff, that these questions address the primary areas of Staff concern and no further questions are anticipated. SCE&G considers the attached responses, together with the previously filed reports, sufficient information for completion of Staff review of our program thus far. We request an expeditious completion of your review and concurrence in our schedule for program completion. Yo rs ver-trul O. W.

xon, NEC:OWD/fjc cc:

V. C. Summer T. C. Nichols, Jr./O. W. Dixon, Jr. E. H. Crews, Jr. E. C. Roberts H. N. Cyrus J. P. O'Reilly. Group / General Managers O. S. Bradham R. B. Clary C. A. Price h. R. Koon C. L. Ligon (NSRC) G. J. Braddick J. C. Miller J. L. Skolds C. Chen S. S. Alexander R. R. Mahan J. A. Blume M. R. Sommerville J. B. Knotts, Jr. NPCF File (Lic./Eng.)

h Questier. 1. Considering (a) propagation path, (b) soil amplification, (c) soil-structure interaction and (d) scattering, what biases in the relative responses of the foundation and free-field sites could be introduced by the shallowness of the explosion sources, compared to the average depth of reservoir - induced earthquakes? (Refer to Joyner's letter). Justify or modify the' estimated reduction factors in light of your answer. Ref erring to the April 8, 1983 letter f rom Dr. W.B. Joyner to Dr. J.L. King, the question as to the effect of source depth on the relative responses of the foundation and free-field sites is stated on page 1 of that letter as follows: "If we accept the description of the dominant portion of the seismograms as a combination of S-waves and higher mode surf ace waves, we have to presume that the relative excitation of the various components of this combination is sensitive to the depth of the source, and f urther, that the relative response of foundation and free-field sites is sensitive to the relative excitation of the components." While the first presumption in this statement is valid, it does not follow that "the relative response of foundation and f ree-field sites is sensitive to the relative excitation of the components" of the S and higher-mode group. The composition of the S and higher-mode group differs f or Test 1, 3, and 4 records because of different epicentral distances, propagation path ef fects, and excitation at the shotpoints; yet Auxiliary Building / free-field spectral modulus ratios are similar f or the three tests. The excitation of higher mode surface waves relative to that of body waves was significantly greater for Test 3 than for Test 4, as is apparent in comparing the records shown in Figures IV. A.1 - 3 and IV. A.4 - 6 of Appendix B. There are several lines of evidence relevant to this issue which are addressed below: (A) explosion tests with distinctly diff erent signatures, (B) wave-forms of RIS events of differing focal depths, and (C) comparison of explosion and RIS recorde. In Section D below, these are summarized and conclusions are drawn for each potential influence separately. 1.1

o A. f.XPLOSION TEST 5 RESULTS Dis inct differences in Test 5 signals generated by shallow shots in saprolite a:d deeper shots in rock provide an opportunity to compare for these two cases the relative response of foundation and free-field sites. Differences in excitation and partitioning of energy at the saprolite/ bedrock interface, rather than differences in shot depth per se, are primarily responsible for the dissimilarity of the records for shots in saprolite and rock. Three of the Test 5 shot holes were drilled to depths 50 f t below the bedrock surface of Charlotte Belt Gneiss, while a fourth (hole #5A) was drilled to 100 ft below bedrock. The subsurface depths at which the initial set of shots was fired were 108, 116, 157 and 210 ft (see Table III. A.2, Appendix B, p.15). Subsequent charges emplaced on rubble at shallower depths included three shots (3, 7, and 8) in saprolite. The Test 5 shots were recorded at a radial array of sites including the foundation of the Fairfield Pumped Storage Facility (hydroplant) at a distance of 1930 ft and a small aperture array on the dam abutment centered at a distance of 3180 ft (see Figure III.A.5). Records obtained in the Auxiliary Building foundation at a distance of 7500 f t are unusable because of peor signal / noise ratic. For the present purposes the hydroplant serves as a good example of a massive embedded structure. The foundation of the hydroplant is 80 ft below grade. In the following comparisons of shallow and deep shots recorded in the hydroplant and on the dam abutment free-field array, the signal amplitudes are not scaled to account for the closer distance of the hydroplant. The sole purpose of the comparison is to investigate the effect i of the two different explosion source types on the relative response of foundation and free-field sites. The dif ferences between the signals generated by shots in saprolite and rock are illustrated in Figures 1.1 - 12. Hydroplant signals and spectra fer shots 6 and 8 are shown in Figures 1.1 - 3. Shot 8 was at a depth of 50 ft in saprolite and 16 feet above bedrock, while shot 6 was 100 feet below the 1.2

7 saprolite-bedrock interface at a depth of 210 f t, quite close to the computed focal depth and hypocenter of the October 16, 1979 earthquake. There are readily apparent differences in the signatures and the spectra, with the shallow source producing relatively stronger high-frequency content than the deep shot. Differences of the same kind are observed f or recordings on saprolite at the dam abutment (Figures 1.4 - 6). Signals from all Test 5 shots form two distinct groups, depending on emplace-ment of the source in saprolite or bedrock. Signatures of the saprolite shots are very similar to each other (Figures 1.7, 1.8, and 1.9), as are those for the shots in rock (Figures 1.10, 1.11, and 1.12). Within each group, the effects of differences in shot depth and distance from the sapro-lite / rock interface are relatively insignificant. The dif ferences between the two groops are attributable to partitioning of energy into different body wave phases or surface wave modes, and to differences in coupling. The dif ference in coupling is evidenced by the similarity of signal amplitudes despite a considerable difference in charge weight (13.5 lb for the saprolite shots versus 121 - 122 lb for the shots in rock). Despite the differences in their signatures, the shots in saprolite and rock produced nearly the same relative response of hydrcplant foundation and free-field sites. Spectral modulus ratios were computed using the three f ree-field stations in the dam-abutment array (stations P1, P4, and PS). No adjustment was made for the difference in epicentral distance (1930 ft to the hydroplant versus 3180 f t to the center of the dam abutment array). Spectral modulus ratios for the shots in saprolite (shots 3, 7, and 8) shown in Figures 1.13,15, and 17 are very similar to those for the shots in rock (shots 4, 5, and 6; Figures 1.14, 16, and 18) for vertical, radial, and transverse compo-nents. In summary, there are strong dif f erences in source excitation and wave character for Test 5 shots in saprolite and in rock, but at most these only weakly affect the -computed spectral modulus ratios between foundation and free-field sites. The dissimilarity of the saprolite and rock shot records is 1.3

not due primarily to differences in focal depth per se, but rather to differ-rences in. excitation and partitioning of energy at the saprolite/ bedrock interface. B. SHALLOW RIS RESULTS To investigate the ef fects of focal depth on the composition of the S and higher mode surface wave group, USGS accelerograph recordings at the dam abutment can be compared for nearby RIS events spanning a range of more than 1 km in focal depth. Corrected USGS accelerograph records for thirteen RIS events are given in Applicant's Additional Seismic Testimony,1981, Volume 2. Velocity records of the RIS events can in turn be compared with the dam abutment records of the Test 5 shots in rock, which were located close to the i hypocenter of the October 16, 1979 RIS earthquake, a little less than 1 km northwest of the dam abutment. Records of a group of three RIS events all located approxitately 1 km to the west of the dam abutment, but with different hypocentral depths, are shown in Figures 1.19 - 27. The hypocentral depths of the three events are 0.36 km (05:47 UTC, October 17, 1978), 1.16 km (08:54 UTC, October 7, 1979), and 1.34 km (23:20 UTC, October 8, 1979). There is no apparent tendency for the records to become simpler and shorter in duration with increasing focal depth, l as might be. expected in the case of simple geologic structure. Comparing l these records with the October 16, 1979 records (Figures 1.28 - 30), it is apparent that there are no systematic effects of focal depth on the records at the dam abutment, as all of these seismograms, while complex, have the same basic nature. Finally, consider the records shown in Figures 1.31 - 33 for a i relatively deep RIS event at short epicentral distance (depth 1.74 km and epicentral = distance 0.35 km). This event at 16:14 UTC on October 25, 1978, occurred almost directly under the dam abutment and yet the records are as complex as any. of the others. These results corroborate the inference that the complex. nature of the waves recorded in this area arises largely from heterogeneities along the propagation path and from near-receiver effects, and that neither source depth nor angle of incidence appreciably affects j ' the character of the recorded ground motion. 1.4 l -~.

~ _. _ _ _. _ Other data for RIS events at Monticello Reservoir contained in a M.S. thesis by Hutchenson (1982)* further demonstrates that source depth does not cause a systematic change in the character of ground motion recordings. In parti-cular, examination of the large suite of digitally recorded seismograms included in the Appendix of Hutchenson's thesis shows: 1. There is little variation in event waveforms from comparable magnitude events that occur in local clusters regardless of depth; this indicates that depth of focus does not introduce systematic changes in duration or wave character of ground motion. Note that this conclusion is based on hypocentral depths ranging from.05 to 1.87 km, estimated from the entire 10 station network (eventa denoted by TN in Hutchenson's Table 3), which gives the best depth determination. 2. An obvious control on the event waveforms is magnit.ude, because rela-tively more low frequency energy is generated for the larger events as the corner frequency shif ts to lower values; this effect is evident in the seismograms presented by Hutchenson. However, this spectral shaping would have no effect on foundation to free-field spectral modulus ratios. 3. There are 'large variations among signatures for comparable magnitude events located at approximately the same distance from a particular receiver, but at a different azimuth; this provides further corroboration that propagation path is a dominant factor in controlling the duration and character of ground motion for shallow RIS events at Monticello Reservoir. i l Therefore, the data presented by Hutchenson further substantiates the fact that propagation path and receiver site effects significantly influence the ground motion observed, while no systematic depth effects can be discerned. Clearly depth of focus does not strongly influence the waveform character of ground motion generated by shallow RIS events at Monticello Reservoir.

• Hutchenson, K.D.,

1982, Source Studies of Reservoir Induced Earthquakes at Monticello Reservoir, South Carolina, M.S.

Thesis, U.

of South Carolina, 120 p. and Appendices. 1.5 l

C. COMPARISON OF EXPLOSION AND RIS RESULTS It is instructive to compare the USGS dam abutment records of RIS with the explosion seismograms produced at the dam abutment during Test 5 (Figures 1.34 36). To compare the explosion seismograms with the USGS records, refer to the center (velocity) trace in Figures 1.19-33; note that the time scales are different. Considering that the accelerograms shown in the earlier figures probably all begin with the S-wave group, the durations and character i of the das abutment explosion seismogramc in Figures 1.34-36 (the traces labelled P2) are comparable with thos'e for the RIS events. This indicates that the same basic wave groups travel f rom source to receiver whether the source is an explosion or an earthquake. As pointed out in Appendix B, because these waves have such high group velocities, they must be body and i higher-mode surf ace waves whose energy travels from source to receiver mainly in the bedrock under the saprolite. D.

SUMMARY

AND CONCLUSIONS The foregoing evidence indicates that the focal depth of the seismic source does not affect significantly the foundation / free-field spectral modulus ratios. The evidence and arguments are summarized as f ollows. -

Firstly, although the composition of the S and higher-mode surf ace wave group differs for Test 1, 3, and 4 records because of differences in epicentral distance (hence propagation path and angle of incidence differences) and in excitation at the shot points, the Auxiliary Building / free-field spectral modulus ratios are similar for the three tests.

Secondly, distinct seismic source differ-ences observed for Test 5 shots detonated in saprolite and in rock produced little variation in the hydroplant foundation / free-field spectral modulus ratios. Thus, differences in the seismic source have little influence on the reduction factors observed for large foundations. Thirdly, examination of dam abutment recordings of RIS events shows that focal depth is not a significant f actor controlling the waveform character of the records. Hutchenson's (1982) data 'further demonstrate that depth of focus for RIS events does not system-atically affect the duration or waveform character of recorded ground motion. 1.6

~ l Also, Test 5 recordings for explosions at different depths in the bedrock show no systematic depth dependence. The influence of specific f actors stated in the question, namely (a) propaga-tion path, (b) soil amplification, (c) soil structure interaction, and (d) scattering, are evaluated as follows. With regard to (a) propagation path, the foregoing comparisons of earthquake and explosion records at the same recording site reveal no significant differ-ences between the signature of explosions in rock and the shallower RIS events on one hand, and the signatures of shallow and deep RIS events on the other. In all cases the. records are dominated by body waves and higher mode surf ace waves whose energy travels from source to receiver mainly in the bedrock. The complexity of both earthquake and explosion signals is attributable mainly to heterogeneities along the entire propagation path and near-receiver effects. The influence of focal depth on (b) soil amplification effects can be evalu-ated by comparing dam abutment records of RIS events of different depth. Compare, for example, the records for an event of hypocentral depth 0.36 km and epicentral distance 1.04 km (Figures 1.19 - 21) with those for an event of hypocentral depth 1.74 km and epicentral distance 0.35 km (Figures 1.31 - 33). The waveforms for these events are similar despite the dif ference in body wave angles of incidence. Differences in soil amplification would be most pro-nounced on the displacement records, but such are not apparent in the data. Two sets of evidence indicate that (c) soil-structure interaction and (d) scattering are not sensitive to variations in the composition of seismic signals at the Summer site. Associated with the recordings of Test 1, 3, and 4 shots are differences in body-wave angle of incidence and in the composition of the S and higher-mode surface wave group, due to different distances to the VCSNS structures (14,000; 3,700; and 4,300 f t respectively). In addition, there are dif ferences in seismic excitation produced at the three shot loca-tions. The excitation of higher-mode surface waves relative to that of body waves was significantly greater for Test 3 than for Test 4, as is apparent in 1.7

comparing Figures IV.A.-3 and IV.A.4 - 6 of Appendix B. Despite these differ-ences, the Auxiliary Building / free-field spectral modulus ratios are similar for the three tests. The second set of evidence consists of the hydroplant and free-field recordings of the two distinctly different source types produced by Test 5 shots in saprolite and in rock. Despite the differences in their signatures, the shots in saprolite and rock produced nearly the same relative response of hydroplant foundation and free-field sites. Considering propagation path, soil amplification, soil structure interaction and scattering, it is concluded from analyzing ear:hquake and explosion records that the shallowness of the explosion sources does not introduce bias in the relative responses of foundation and free-field sites. There is no signficant difference in the spectral modulus ratios observed for different S+ higher mode levels of excitation and different angles of incidence. Therefore, the explosion tests are appropriate for estimating foundation to f ree-field spectral response for shallow RIS events at Monticello Reservoir. I i 4 1.8

9 D a e o (n Q r 3 L 4 o (O L t a _oa g s tn en w ,, S C Y o O %s aQ L z g j Y Z E).c c ]% N.* D = u 'C O o o ~< m (n a> u.- h 38- ~ 0 1 oI ar oe me ,L 8,a 's mW C (D E .o,o./ = oo w Hg / .-. (n 8m t / .=C b-- e $witi lni, a Ln g g*@ !!tti t i i o i m; 4.s ~ m o m n o g.o u p-c- g-to o e o e o o C (3 o o o o o o o d d d d d gg 5 i e i Bad u [3 zo8 b h m~ w d 4 13 ~ (f") d b I z ez y 9 'D C 4 W o e l CD d oo ~ b g=b _CC H l / Ln ln ) w E d 44 b w J 9 h,@mu W / C cn v g D <.'f j e ~ e O h m ~ o l ~ en t ~/, Q / OI e d NC s / w m ha f CD h ~ O ,/ CO - en H O '! t ! ! !!!!!! i i hW L1_ s I { ] t 1 l t _o I 1 m -w e o o o -m e C Q O o e o o o o o o o o o o 1 a a o o o o o

T T \\ T 1 tn = Y ! /u j'h/ V'g/ 'i I / \\ 4 I \\ l\\ ~j \\ \\ ~ I f I r \\j'\\ \\!\\\\ "\\' l T Al T [i 3[,\\l \\

tl I

I I I I I l l l l l

0. 0 10 20 30 40 50

0. 0 10 20 30 40 50 BS.S08.RHPS BS. SOB.RHPS BS.S06.RHPS BS. SOS.RHPS 0 002 o.003 o.002 0.001 0G h

,fy^y'= $Ue?l~+"'kyMi ~: ' " l o, o m Efg,LfPf;2.A4's.Je2 ;u,J. A i -0.001 j -0.002 - -0.001 -o.003 1 I I I I I I l l l 1 l

0. 0

0. 5

1. 0

1. s

2. 0

2. 5

o. o

0. 5

1. 0

1. 5

2. 0

2. 5 PLOT NO 6

HUN NO 507 PLOT NO 4 RUN NO 507 FIGURE 1.2 Radial ccuponent seis:rograms and Fourier spectra for Test 5, shot 8 (50 ft depth; left) and shot 6 (210 ft depth; right) recorded on the hydroplant foundation (IIP).

= i 5 e-i 3 A g m l\\ i }\\ ? t n =! / /k / 1/ \\ M I /\\i/ l\\ I t / -I \\/ j \\r\\^ j \\fr n,\\ ? IMI w al 5 j% \\ y w u 1 I I I I ~ i i i i 1

0. 0 10 20 30 40 50

0. 0 10 20 30 40 50 85.S08.RHPS B5.S08.RHPS B5.S06.RHPS B5.S06.RHPS 0.002 0.001 0.001 5.0E-4 l

0. O %

l } d i h. nnan.nn.n.s. s r - an. 0* G j [f W [Y " "'vm 7T "r " "'n ' I i -0.001 l -5.OE-4 -0.001 -0.002 -0.002 - 1 I I I I L L i I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5

0.,0

0. 5

1. 0

1. 5

2. 0

2. 5 s

PLOT H0 G RUN NO 508 PLOT NO 4 RUN NO 508 FIGURE 1.3 Transverse cmponent seismogrcus and Fourier spectra for Test 5, shot 8 (50 f t depth; left) and shot 6 (210 ft decth; right) recorded on the hydroplant foundation (IIP),

n. ~ ,c ~ ~ - ~ _ r ~fNd i r l m E L T M

• \\^Iv\\,\\;i\\p 1

g 7, lW w ~^ ) 4 [ A f lu / g 5 r i g jg i La \\l b 9 !~ \\,\\.1 l i I I I E I I I I I

0. 0 to 20 30 40 d0

0. 0 10 20 30 40 50 5.308.RP5S 85.S08.RP5S B5.S06.RPSS B5.S06.RPSS 0.00G 0.004 0.004 t

j 0.002 [ r. O.002 ~ l h

0. 0 O. O 7

-0.CO2 f j -0.002 ) -0.004 -0,004 -0.006 -0.008 ~* b I I I I I I I I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 LOT NO 6

RUN NO 509 PLOT NO 4 RUN NO 509 FIGURE 1.4 Vertical component seismograms and Fourier spectra for Test 5, shot 8 (50 ft depth; Ieft) and shot 6 (210 ft depth; right) recorded in the free field on the dam abutment (PS).

,e I e n =- i 5 E E n p 6 /\\ \\r\\vh / \\.. u w ]. ' \\ +. <l

r..

5 ,\\ .. r' \\ i

• l

) V 'I ,e i e-p ~ V i a T r s 5 I I_a . _i I 7r l l l l l

0. 0 10 20 30 40 50

0. 0 10 20 30 40 50 B5.S08.RP5S B5.S08.RP5S B5.S06.RP5S B5.S06.RP55 0.006 O.004 0.004 f

0.002 fl 0.002 lI i

0. 0 m f

[ bnA

0. 0

-d v'y yv^ & a g { g-yvv v-q -0.002 {l l [ g -0.002 { -0.004 -0.006 -0.004 1 I I I I I I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 PLOT NO 6

RUN NO 510 PLOT NO 4 RUN NO 510 FIGURE 1.5 Radial canponent seismograms and Fourier spectra for Test 5, shot 8 (50 ft depth; lef t) and shot 6 (210 ft depth; right) recorded in the free field on the dam abutment (PS).

v v =- i = 4 2 m m .\\ 6.' \\ \\ .\\,. s : ' ' \\.f 1,, f^Y "yd le \\ / e,,. ti 't 9 'i.\\. /, o m e = E I I I i 1 l l l l [

0. 0 10 20 30 40 50

0. 0 10 20 30 40 50 85.508.RP55 85.S08.RPSS B5.S06.RP5S B5.S06.RPSS l'

0.004 O.002 l l 0.001 f 0.002 p k j l } j y gu'^v

0. 0 e j

- -h Y--

0. 0 i

-0. 002 5 I -0.001 \\ h j -0.004 l) -0.002 -0.00G -0.003 I I I I I I I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5

" LOT NO G RUN NO 511 PLOT NO 4 RUN NO 511 FIGURE 1.6 Transverse couponent seismograms and Fourier spectra for Test 5, shot 8 (50 ft depth; left) and shot 6 (210 ft depth; right) recorded in the free field on the dam abutment (PS).

BS.S03.RHPS BS.S03.RHPS ~ 0.002 - O.001 0, 0

[yYM.$.W.'-,N n'-i- % "

-0.001 I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 BS.S07.RHPS BS.S07.RHPS 0.006 -

O.004 D.002 O. 0 'b,f,'y,;, '; '; ';: -0.002- -0.004 I I I I I l

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 BS.S08.RHPS BS.S08.RHPS 0.000 D.006 D.004 0.002 -

I O. O b/-y^ h'.'i.. -0.002 - -0.004 - -0.006 -0.008 - 'l 1 1 I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGURE 1.7 Vertical cmponent hydroplant foundation records of Test 5 shots in saprolite. Tne subsurface depths of shots 3, 7, and 8, shcrm frcm top to bottcm, were 46, 67, and 50 ft.

B5.S03.RHPS BS.S03.RHPS I 0.001 l , k k d A ut, t..i a o. i i.1-

0. 0 g p yrig m prr' v r'T ~'ri-r -

-0,001 I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 BS.S07.RHPS B5.S07.RHPS 0.002 O. 001

^J)/f,'.',8,",,y4j,.';J'"'i'I^E#'?U O. 0 f r -0.002 -0.002-1 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 BS.S08.RHPS BS.S08.RHPS 0.003 O.002 O.001 O. 0 --'

' ('g- '-i,;fYc[,','ls"Off/.'c*r, -0.001 -0.002- -0.003 1 I I I I I O. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGURE 1.8 Radial cmponent hydroplant foundation records of Test 5 shots in saprolite. The subsurface depths of shots 3, 7, and 8, shom frcrn top to bottcrn, were 46, 67, and 50 ft.

BS.S03.RHPS B5.S03.RHPS 0.001

,'l,\\j-Y}l? '.- i'.,7 v ' Y

0. 0

^ J b I l -0.001 1 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 BS.S07.RHPS B5.S07.RHPS 0.001 f 'NbEA

'[,"'p"n^2,Ul[7 O. O i, -0.001 1 i i I i 1

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 85.S08.RHPS B5.S08.RHPS 0.002 0.001 b (Nllll,'O ?.','$,;. ':R'-

I'

0. 0

-^;, -0.001 -0.002-I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGURE 1.9 Transverse cmponent hydroplant foundation records of Test 5 shots in saprolite. The subsurface depths of shots 3, 7, and 8, shown frcrn top to bottcm, were 46, 67, and 50 ft.

B5.SO4.RHPS B5.SO4.RHPS 0.004 I 0.002 b O. 0 jffy%,' ^Y *:-'

• : ~=

-0.002 -0.004 -0.005 - I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 B5.S05.RHPS B5.S05.RHPS 0.002 O.001

0. 0

-{?,i^wh; y ^',^,=- - ^ A ^- = - -0.001 -G.002 - -0. 003 - -0.004 - I I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 85.S06.RHPS B5.S06.RHPS 0.003 O.002 l

0.001 f);g^. " rf- ~, -

0. 0

-0.001 -0,002 - -0.003 - 1 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGUPI 1.10 Vertical ccmponent hydroplant foundation records of Test 5 shots in rock. Tne subsurface depths of shots 4, 5, and 6, shov:n frce top to bottcm, were 116, 157, and 219 ft.

i B5.SO4.RHPS B5.SO4.RHPS 0.002 0.001 l

0. 0 f

[,yyT y/ 7 // f (A,A f _- A -_',* -0.001 I -0.002 -0. 003 - l 1 1 I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 85.S05.RHPS B5.S05.RHPS 0.001 O. 0 =

,y'7k'dp4?f,"#fe',-# -0.001 1 1 I I I 1

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 B5.S06.RHPS BS.S06.RHPS

0. 002 0.001 I

i I V MI[ red '* YON # ^# 'd'

0. O "

-0.001 1 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGURE 1.11 Radial canponent hydroplant foundation records of Test 5 shots in rock. The subsurface depths of shots 4, 5, and 6, shcun fran top to bottan were 116, 157, and 210 ft.

B5.SO4.RHPS B5.SO4.RHPS 0.002 O.001

0. 0 jyfl Jay /p[wfy",w,7 j

-0.001 1 -0.002 1 I I I I i

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 B5.S05.RHPS B5.S05.RHPS 0.001 5.OE-4

[g,'M@,y,YAc1'f q 0, 0 2 -5. GE-4 l l -0.001 1 1 I I I I

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 l

85.S06.RHPS B5.S06.RHPS 0.001 5.0E-4 9 OpfAwmwe.=

0. 0 =,

g -5.OE-4 I -0.001 -0.002 1 I I I I 1

0. 0

0. 5

1. 0

1. 5

2. 0

2. 5 FIGURE 1.12 Transverse canponent hydroplant foundation records of Test 5 shots in rock. The subsurface depths of shots 4, 5, and 6, shown frcm top to bottcm, were 116, 157,and 210 ft.

o D ^ cl I I I\\ l\\ \\ l r I tn I \\ I\\ \\ ,\\ I 'I d l \\ l\\ \\ j ~ \\ I \\ I \\ \\j \\ I l \\ I \\ \\; \\ I o I \\ \\ \\. \\ I I / \\ l \\ \\ -~ I I \\ \\ I l \\ I I I \\ l \\ \\ I tn I \\ I \\ \\ I l I \\ I \\ \\ I I l \\ I \\ \\ ~ I^ l \\ / \\ I

\\

\\ /\\ \\ )I \\ o I \\ I I\\ \\ / \\ j\\ \\ I I ___p1_ L d _ $, l_ _ _ _ J _ _f_\\ I 7 j \\ II \\ l \\ j \\ ~ \\ \\./ / \\ j \\j/ \\ / \\ j v/ \\/ \\ j \\ _ _,/ tn I /\\ c / / s, \\, O N / / </ _ o o 10 20 30 40 50 TEST 5, SHOTS S,7,8. HYDR 0/Pf(D 98RRY V RATIOS FROM MODULI INDIVIDUALLY RUN NO SD2 POWER SPECTRAL NOISE SUBTRACTED FIGURE 1.13 Vertical ccuponent spectral nodulus ratio, hydroplant foundation / dam abutment free field, for shallow Test 5 shots (3, 7, 8) in saprolite. Note that the signals have not been scaled for epicentral distance (1930 ft to the hydroplant vs. 3180 ft to the center of the dam abut:nent array).

o l,\\ 7-l\\ m I I \\ \\ \\ \\ I I I to I \\ N / l t/ I \\ I \\ / I I \\ z \\ I l ^ \\ l \\ \\ O I \\ l\\ \\ \\ \\ \\ \\ 1 I l\\ \\ \\ I \\ I N I l I\\ \\ 1 I \\ l I \\ l\\ \\ \\ I \\ \\ tn \\ It I \\ \\j \\ \\ I I li I \\ \\ I I \\ l ~ I I n I \\ I \\ I' \\ l \\ f\\ z,

Il - \\

o - -] - l -\\-- - - -- -/T J l \\ l \\ l / l l\\ \\ ~ jy I J \\ l \\ l \\ I y j j y# g s j / \\/ \\/ \\ I \\ / / /\\ / \\ \\ / sNN/ / \\/ \\/ o j L O l .o 10 20 30 40 50 TEST 5. SHOTS 4.5,6. HYDR 0/PRD RRRRY V RATIOS FROM MODULI INDIVIOURLLY RUN NO SD5 POWER SPECTRAL NOISE SUBTRRCTED FIGURE 1.14 Vertical ccuponent spectral modulus ratio, hydroplant foundation / dam abutment free field, for deep Test 5 shots (4, 5, 6) in gneiss. Note that the signals have not been scaled for epicentral diste.:,ce (1930 ft to the hydroplant vs. 3180 ft to the center of the dam abutrent rrvwL

o m Ii f I \\ l I 1 l n LD I\\ \\ \\ I ~ l\\ l l l na I\\ \\ \\ I I\\ \\ \\ l a I \\ \\ \\ I I \\ \\ . l I N I \\ \\ l} I I l 6 I ) I tn I \\ I\\ ,,\\ I \\ I\\ \\1 \\ I\\ g j \\l \\,, j 4 v \\1 \\, a j

---

7 I-\\

9 I \\ / l \\In ry, 7 ~ j \\/ / \\ / \\ / g / f W /\\ / \\ / N / g / j . /, / \\s/ \\/ \\ I \\.j o /s// N ' \\. sl/ l o o 10 20 30 40 50 i TEST S, SHOTS S,7,8, HYDR 0/PRD RRRRY R RATIOS FROM MODULI INDIVIDURLLY RUN NO SDD l POWER SPECTRAL NOISE SUSTRACTED l l FIGURE 1.15 Padial ccuponent spectral modulus ratio, hydroplant foundation / dam abut 2nent free field, for shallow Test 5 shots (3, 7, 8) in saprolite. Note that the signals have not been scaled for I epicentral distance (1930 ft to the hydroplant vs. 3180 ft to the center of the dam abutsnent array). l

o rm p D I \\ \\; n/ \\ I i l I / \\ I l l m gl / l l t N e / l I I l j I \\ l m I \\ \\ l \\ \\ l \\ f\\ \\ D I \\I j J / \\l / \\/ g/ I v f\\ o / n p j - l/-\\ - - -4 7 t-- j \\ l \\ ~ / \\

\\

/ \\ / \\ \\ I / \\/ \\/\\ J / m j / j .o t' Il \\ e4 c._ p - o o 10 20 30 40 50 HYDRO'PRD RRRRY R TEST 5, SHOTS 4.5,6. / RATIOS FROM MODULI INDIVIDUALLY RUN NO SDS POWER SPECTRRL NOISE SUBTRACTED FIGURE 1.16 Radial caponent spectral nodulus ratio, hydroplant foundation / dam abutment free field, for deep Test 5 shots (4, 5, 6) in gneiss. Note that the signals have not been scaled for epicentral distance (1930 ft to the hydroplant vs. 3180 ft to the center of the dam abutment array).

e o en tn ("\\) i I N l / 'g o I e /s \\ / \\ / csi gf v p I \\I

\\.

I \\) tn i g / \\ ~ j l i / n I /\\ / \\ o [ i _l_ \\ _/ __ _ __ _ _ q \\l 7\\ / ~ \\ / \\ / /T I w j/ j\\ \\./ \\ /\\ I h / \\ \\ e s/ \\/ \\ /' \\ // g / \\j \\, o / j \\, / v -s e/ ~ _. /, -I I I l O o 10 20 30 40 50 l l TEST S, SHOTS 3.7,8. HYDR 0/PRD RRRRY T RATIOS FROM MODULI INDIVIDUALLY RUN NO 501 POWER SPECTRAL NOISE SUBTRACTED i l FIGURE 1.17 Transverse ccnponent spectral modulus ratio, hydroplant foundation / dam abutment free field, for shallow Test 5 shots (3, 7, 8) in saprolite. Note that the signals have not been scaled for epicentral distance (1930 ft to the l hydroplant vs. 3180 ft to the center of the dam abutanent array).

o co w l x j d I \\ I r\\ l 5 I j\\ \\ l 7 O \\ l \\ \\ N l \\ \\ / / J \\ \\/ f l \\ / I \\ w l p\\ g l I y I l\\ I \\_ _q _ l h I I \\ n \\ l \\ \\ r\\ ~ / \\ ll \\\\/ t \\ / /l / /'g \\ II \\v / \\ / \\ LD // / \\ l ( _, / \\ \\ y \\,7 1 N /i I j \\, J (Q'/ / o l d 10 20 30 40 50 l lEST 5, SHOTS 4.5,6. HYDR 0/PRD RRRRY T RATIOS FROM MODULI INDIVIDURLLY RUN NO 504 POWER SPECTRRL NOISE SUBTRACTED FIGURE

1. 18 Transverse ccuponent spectral modulus ratio, hydroplant foundation / dam abutanent free field, for deep Test 5 shots (4, 5, 6) in gneiss. Note that the signals have not been scaled for epicentral distance (1930 ft to the hydroplant vs.

l 3180 ft to the center of the dam abutment array).

ll

ll i

t I s 6* >dd cddb;b "r 6td 6U$Wu t P E 0 ~ / R K 'h V A F I L I 4 i I U G U I l E f S R l E I J R C 1 C 1 E 9 L-5 dMS 7 eLo p u 5 t2t h. h F 2 IC 0 c LN Ru T/ 3Ii ES 6Sg 1 RE ken EC mve D/ en S ant FE nt T d c RC l O. oo H efr HV E p r E i0e L c5c 2O e: t . C n4e S t7d 0I E r 0T aLU C T 0Y lI S O d,C G N T S O i O-S s1d 0 t0a n/m a 54 1 c7a 09 e/b7u 0C 1,8 t 0N m H/ 0 e 4wn ZS it E kt 2 C .m h r e 0 hc yo pr S od P c L eo nf trt 0 w~M ah le 0 0 7 C

i 5 g 1rg $E* - n2r ' = ~, n R - s M P g "! g ~ 7 o 5 m e u P a' C eN O3 s T e . C ma nE=

• .E a

e,as ~ N OC I E u <L T S A/ RC EE RY LS E / C M CC A T T C u I E CS O/ L M V . [ EC TN E M E C M I A C L P S I D e H u 3l O3. ? g= ",8 8$c@ I g.g= aO c5 B ".P, g

• 0 Q.' e $ e n~. p "z o

".m b akb $ O? 4S cEa$ r O) o. M

'i Ii! 1; a oG t> Euuw w Te 8 o,u.n_e *rw2s >(L_ a a or o r s y' u ursvn uwu sa s t P E A 0 ~ K f4 V F A I L G U U E RE S A 1 C 0 2 C 1 E L-e4a d E 1 3 p s 2 t.2 t h F 2c 0 u IC R LH 3Iq T/ 6S0 ES n RE ke e EC mvn 1 et D/ an S ntc FE d o RC T or O. efr l p e H HV i0c E E c5t L e: e 2O n4d . C t7 r U S 0I 0T aUS E lI G C 0T d,C S O = i d N T s1a O O-n/m t0 S 1 a 1a 53 c7b 02 e/u 7t 18m 0C .,e 0H 0 n H/ 4wt ZS i l E atr C .n h ec ho D yr 2 I pd S c P co L ef n tt 0 rh ae l 0 W 1 7 C mH

e e n O e a ik 4 ,,P 9 ay ly= g,

• e sa "a e

5E g r~8 L

g"g a = C r o o LC H AI wO n. n.f J G o uf g l11 5 EE" 8ts .C ds 2 QDh OE bO g m2 m tyr 8 \\sa m )/( A O ~ e m* FF H -+ D 3 -+ e 2 Q x q m 4 CES/CES/MC CES/MC MC NOITARELECCR YTICOLEV TNEMECALPSID

+ O 6 e 0 x< g F wn m7 mw*r" ';: C m=sMa.< o2 <' P 8 nxN4m *g yr' 9 m ? O o nK u - = L m = :r N*0oo 4 $ ooZW F N OC I E icM T S R / RC EE LS E/ C M CC A -+ h T O / D TC I E CS LM ,J EC V Wc T N E M E C M AC L P S I D O ~ m s, nE=0 msl'

• .U dnM'?

$ " 8 4 s"a C g g,3 Pr'T 8r r na om e5 ) O3 ". M F $ ns. ' e m6 ym Sg" a mh - o

• , $c Gg

' j ? CE E h ~.0 E L

O' 8 2n b C. - o 7 EE 4 s h Rj a,0- ~ Mhf 35 20 n& no

• P l5F 8m R$

$siE o 0= gsm a = n aC 4-o AC d 5 ~g n* I am 83 r mE[L

== Y8 R PE nam mm

== h,g m .g @E mm

• C D 3U A

& '5 L s w m

• .h z

h lT't n 2 ne m C a o A Il1* +- CES/CES/NC CES/MC MC NOITARELECCR TTICOLEV TNEMECALPSID

I r . ~ = H = 3 C 3 1 0 el ha 0 t r t L f n oe P c S do rp I D oy ch e C rhm E tk SZ ti /H nw0 e N0 4 m,1 C0 t 9 u7 00 b/e 5S 2 a7 c S 0 n 0 O m/a a0 t O N d1 s T O i C S , d T0 E GC 10 S SI l UUa 10 r C. d4t O2 e5n L E t : e E H c8 c e0 i YM I r p .O I rf e CR oo EF c d S tn tna /D ne CE evm ER nek SE o /T pS 6 NL uI1 CI .1 uR c 1 F 7 h 8. h t2t 3 u p 6 o Le 1 Sf d L E 5 C C 2 A 1 j S E E R U U L I G A j \\ l I V F I } 0 N A ~ E P t i;" 5 d 5 h MG Es* 5E!mb Il

,l; !lill;! l'l!i!. !l yOUce:N c >oaC

• - &5 5b 5

u us5us u t P E O y( R ~ F M IG 4 U I V H R J I g \\ L E i U V l \\ E 1 1 I S 2 6 A C C E L e4eV d p r h.2 t t i 3 8c 0 1 a Rl 0 1I F 6Sc 1 IC w LM kemvW T/ e ES ann RE nte EC d n ot D/ ef S p c FE 1 i0o RC 1 c8r O. e: r M MV n5e E E t 4c L r t aUe 2O I d . C l d,C S 0l U E 01 i S C 0T s1G O t 0S = N a/ 1 n0d U 0-7a S c/m 0 e 7 54 1,9 a 2 b 00 4 u 0w t 0C im 0H kte H/ mhn ZS . h.t E yr ,C peoc D co er I nd S t P ro L af l t h 0 e 0 1 0 C M 3

w ~ m H =, C 9 l 2 ea 0 hr tt 0 n f e oc o L dp P ry S oh I c ehm D. rtk i C t w5 E n 0 SE e,. n91 /H 3t 7 H0 u/e C0 b7c a0n T0. Q m/a 50 0t 35 a1s 9 d i S ,d 00 SC D 1 GTl N SUa O U r C 4t T0 E d5n t0 e:e S t8c c. e2 c0i e p l rfe E E ro VH M o d .O ctn I CR na T EF tenvm S eek /D n CE cS6 ER I1 qR SE w 8. h 1 /T HL CI t2t F s p 0 a Le EMd 7 6 7 L 2 E 1 C C R ER U S G E I U F L l k R V 4l 4 M R h ~ E P t SmDwUru uom ro ru t u sO> z"J Uaan.* a zS e@w wuuG .L i;:1i < j -j ,1

<l,i

= m ~ = M - 3 C C l a T 6 r U0 t en 6 0 t c he 0 o 76= f p oy L h 0E d P U rh S ot I 0 ci 8 ew. D r m 1 -. ,k 9 .C t9 E n7 8 7cSf e 9u/H m/7 T 6 H0 t1 0 G u/ 1 C0 b0e 0 a1c 7 60 n 0 6.45 = - 2 m,a aCt 19 S dI 3 Us 7 D i 0 / N S d G6 R6=1 O S0 C U:l E1Y0 a / E 7r B0T0 S d0 t I0 /D t e e e n O1.C = OI chc TN et i L A E r p C E M rf D e O VH oo I .O c d O T tn CR L t na FLEF ne S evm OE nek C CE o I ER gS7 T E SE mI0 N xR KO /T RM c 0 NL h 8. h .C I . I F t2t U .9 u ,p C o Ie O .6 SMd HS 4 T .3 RE AL= 8 L E 2 L EI C 1 .V C S A ON E I LK: R S U LN G E E I E U F J kI L C I A 1 I V I I g T K 0 N A ~ E O P M t uNsa e5 oe5 ru $fUcc &R b Nee 61 C gd> i c k ,i !i .ll l 1i i l

MONTICELLO EARTHOURKE OF OCTOBER 16.1979 -.0706UTC JENKINSVILLE. S.C. MONTICELLO DAH.IO/16/79.070GUTC.CONP UP t PEAK VALUES: ACCEL = 175.5 CN/SEC/SEC. VELOCITY = 1.556 (M/SEC.DISPL= -0.03 CH FILTERED FROM-1.000 TO 50.00Hz I i l z?e ~ \\l j 0 f rr C g g 4 ~ ~ et i J ) uu - S f I E: ge ~~-~ _ _ d5 f 0 j l t I 1 8 i Z y M e5 3 E' \\ N ~ O o i i U . l' 1 2 3 l TIME - SECONDS ( FICURE 1.29 Vertical ocroponent corrected USGS dam abutment record of the { M,2.8 RIS event of the 07:06 UIC, 10/16/79, with hypocentral Idepth 0.07 km and epicentral distance 0.78 km.

NONTICELLbEARTHOURKEOFOCTOBER 16,1979 0706UTC JENKINSVILLE. S.C. HONTICELLO DAH. 10/16/79.0706UTC.90 DEG t PEAK VALUES: ACCEL = -350. CH/SEC/SEC. VELOCITY = 3.079 CH/SEC.DISPL- -0.05 CM ~ FILTERED FROM l.000 10 50.00HE bu j ~

ti Eu 4-u-

= Ur yu 1 l l ~ -su l l ( i t z i E f h3 l d E. ] 1 0 1 1 2 3 TlHE - SECONDS FIGURE 1.30 East conponent corrected USGS dam abutment record of the ML 2.0 RIS event of the 07:06 UIC, 10/16/79, with hypocentral depth 0.07 km and epicmtral distance 0.78 km.

i 1i 'r ,] i I, ^

D ah w5 w0u 5 g0>*cG 8z EL c

r uNEo u2YoCs"u t P E ^ 0 o A / N t T F A V A I L G U U 4 R E l \\ C S L A 1 C 3 C 1 E L-dMS 7 eLo p u 0 th.2 th 5 2 F c 1 IC 1 Ru LM 7Iu T/ 4 Sp ES o ken RE mve EC en D/ ant S nt FE d c oo T. RC efr O. I p r N HV i1e E E c6c L e: t 2O n1e . C t4d S r 0I aUU E 0T lI S C 0T d,C G O = S N T is1d O O-t0a S 0 a/m n2 55 c5a 2 03 e/b 7u 0C 0,8 t e m 0H 2 3 e H/ 5wn ZS it kt E C

m. h're hc D

yo I pr _e S od P c L eo nf t rt 0 ah le 0 ~- 1 0 N. = C 3 N

. i; t PERM VALUES: ACCEL- -37.5 CN/SEC/SEC. VELOCITY- 0.6109 CM/SEC.DISPL- 0.015 CH ~ ~' FILTERED FROH 2.000 TO S0.00HZ z ~ .-.U w sn \\ t n uu [ ~f s CM .I sy f f~V tu o I 3 . 'h ~ .4 e 4 5 4 E ~ ~ in ~ ur cc u ~ a 0 1 2 3 l TINE - SECONDS l I FIGURE 1.32 Vertical ccmponent corrected USGS dam abutment record of the M,2.2 RIS event of the 16:14 UIC, 10/25/78, with hypocentral l I depth 1.74 km and epicentral distance 0.35 km. l

? PERN VALUES: ilCCEL- -95.8'CM/SEC/SEC.VELOCliTa 0.50tl CM/SEC.DISPL= 0.010 CM FILTERED FROM 2.000 10- 50.00HZ M ~ l ~_ - a _- __e . ~ - - - d8b m 1 = 1 U Yu oE i i d5 3 l ~ 4 D l 1 E \\ l mb i y uI a l 0 1 2 3 TIME - SECONOS FIGURE 1.33 East uxwnent corrected USGS dam abutment record of the M. 2.2 RIS event of the 16:14 UIC,10/25/78, with hypocentral depth 1.74 km and epicentral distance 0.35 km.

e 5 5 5 5 l 4 4 b he b-km km km E-e$ E-

• N
• N E.

E E E 5. E. E. E. O gno (no (n o (no (n O 1/10 (n D (n D N T <C da. g s o ) ( m )> h ( ( ) l sys* ( us@l 8 er i E ice ) 3 83 ) .{3 5 J L l sn %e t 5> >g f o

• Eg 3

f e b s case > <c <>y< > fb38 >7\\, g.s H ) i O ) I C { 4 > C> <}; O E ": 9B i B }<i C>(==R =- 8 ( c, [ -( $c ffS i ? ~ y l T )< tn J== c: m i w W l & O O d O N tr) v in CD 6 G3 CD

O TEST 5 R

o. So_4 1.00

i. so 2.00 2.so

o. co g

1 I I I AO T5.SO4.RPiS ^ i i p o.oisa a f i A b^b [\\ Ts.SO4.RP2S L o.o s3 j f G " \\/ V \\ b /\\ A A /'\\ m T5.So4.RP4S ^v] 3 V-y yyv y N o.oiss b/ b A [ A A /\\ T5. SO4. RP5S .gl y v y v gj /'7 n1 .g 4 j 1 vv

o. m s3 i

5 W---=-.^#f --- = T5.So,4,,R,FRS I = T5.So4.,RABS l 6

---

w-

-w ~-- -- N [v^w^^ w^, T 5. S,0,4,.,R,oSS 7 8 v'w^^^ N ---- Ts.go,4,.,R,HPS i 9 FIGURE 1.35 Radial conponent seismograms for Test S, Shot 4, detonated in rock at a subsurface depth of 116 ft. 'Ihe acierograph pad trace (P2) is second from the top. 'Ihe epicentral distance to the pad is 3100 ft l (0.94 km). l l

j e e a e 8 A to to th to to to th (l') N T 10 CC (D to 11. kn km hm m m m m m .c .e .c .c .e .c .c .c TO TO TO TO TO To To To o d d d d d d d md 9a o W W W W m W m m gjg* N H H p p p p p y o ) C "55N ) ( ( = o 0 2.s 2 ) ~ l i ) g.ob w ) / \\ u&m e s t a v.s e (> <)< 8 l a.-. 3 Fg J g ) c ) h 8 u-o >P >> <C 2 c a 8s o >) r 5 =".b 5 u "#m" = >) q 5 Eg l 1 t <C \\> ? S$m e @ go 5 ( C 1 3 L 3eg1 l 8 N 4 t go. pc-p e

== l } i 1 \\ 038$3 8 l388 ,e c h s s [s> ><> <a geggs ~ e e i c c 3 b~h S [> ( C }C[ c:(( l l l m p -x o ^ L d 3, q B f7 jr 1 i j 5< c m T l H Lt) H-oo d N m w m D N D o

Question 2. How sensitive are the final response-spectra reduction factors to the assumption of zero-phase shif t for the foundation-to-free-field transfer functions? Justify or modify the estimated reduction factors in light of your answer. The sensitivity of the f oundation response spectra to the assumption of zero phase shif t for the transfer functions is evaluated below by comparing results for three different cases.: zero phase, random phase, and the observed phase for the explosion data. The zero phase and random phase filters cause minimum and maximum dispersion respectively. Response spectra of foundation accelerograms obtained from transfer f unctions with the observed phase are bounded by the zero phase cnd random phase cases. Given below is a description of :he procedure by which the 1979 accelerograms are filtered to simulate motion that would have been recorded at f oundation recording sites. The construction of such a filter f rom the smoothed spectral modulus ratios necessitates an assumption about the phase angle as a function of frequency. Of these three possible models, it will be shown that the assumption of zero phase gives a conservative estimate in the sense that the filtered output amplitude tends to have the highest peak value when zero phace is used, in general, response spectra obtained for the zero phase shif t assumption envelope those obtained by other assumptions about the phase. in particular, the zero phase response spectra always lie above the random-phase response spectra. Figures 2.1 - 4 illustrate how the filtering process is performed. Figure 2.1 is the input ground acceleration (90 degree component) as recorded at the dam abutment during the October 16, 1979 earthquake (maximum acceleration as shown in the plot is about 330 gak). In Figure 2.2 is plotted the smoothed spectral modulus ratio between the transverse component of ground motion as recorded during Test 4, Shot 4 at the Auxiliary Building and the USGS acceler-ograph pad. In this case, transverse is in the same direction (E-W) as the earthquake motion input. Because of poor signal-to-noise ratio beyond 50 Hz, the value of the transfer function is set to unity between 50 Hz and the Nyquist f requency so that the spectral amplitudes of the input seismogram will 2.1

o + I i i be passed with no amplitude distortion in that band. The aim is to construct from this amplitude function a time-domain filter with which one can simulate the ground motions as recorded inside the Auxiliary Building during the 1979 earthquake. To do this, the phase information is nocessary. The assumption used in Appendix B was that this filter has zero phase response (the straight line shown at 1.5 in Figure 2.2). Inverting this -filter to the time domain results in the impulse response given in Figure 2.3, Note that in this plot I the onset time of the impulse response has been displaced from the lef t-hand abscissa for clarity of presentation. The entire respcase appears to consist of a single spike of amplitude 420 units; however, this is purely an artifact of the zero phase assumption, and results only in the unaltered transfer of energy (amplitude or phase) between input and filtered seismograms beyond 50 Hz. The information of true interest is the far smaller oscillatory motion shown in Figure 2.3. When the input (Figure 2.1) is convolved with the -impulse response (Figure 2.3), the result is the " filtered seismogram" shown I in Figure 2.4. Note that the peak amplitude has been reduced from about 330 ) to 95 gals, which is a reduction factor about aqual to the average spectral modulus height in Figure 2.2 between 0 and 50 Hz. 4 .Results for random phase are shown in Figures 2.5 and 2.6. Figure 2.5 shows the filter impulse response which results from using exactly the same spectral modulus as given in Figure 2.2, but now the phase has been taken to be a random number between pi and pi. Figure 2.5 shows that the character of the filter has changed to random oscillations over the whole record. The effect on the resulting filtered seismogram is pronounced. (Compare Figure 2.6 with the zero phase result in Figure 2.4.) Note that now the peak amplitude on the filtered seismogram is only about 35 gals. l To investigate further the ef fect of phase on the simulated seismograms produced, transfer functions using the observed phase are shown in Figures 2.7 an.1 2.8. Figure 2.7 gives the impulse response from a filter which has the same amplitude characteristics as those of Figures 2.3 and 2.5, but now the phase is selected as follows. At the time the spectral modulus ratios are constructed, a complex transfer function is also computed (in this case l 2.2 e m-- e ,~. _.. -m..--~ m- - - - --

• 1 between the pad and the Auxiliary Building).

This transfer function has associated with it a phase angle. The impulse response of this filter is given in Figure 2.7. The resulting seismogram, given in Figure 2.8, is seen to have a peak amplitude intermediate between the zero-and random phase cases (compare Figure 2.8 with Figures 2.4 and 2.6). Figures 2.10 - 2.15 give the filtered seismograms which result when the 180-degree component of the 1979 accelerogram (Figure 2.9) is filtered by a spectral modulus ratio from the radial component as recorded at the Auxiliary Building and the pad. Figures 2.10 - 2.15 compare with Figures 2.3 - 2.8 for the 90-degree component. In summary, filtered seismograms, each constructed 'using the same input data and transfer functions which share identical spectral amplitudes, have the highest amplitudes when the filter employed has zero phase. Figures 2.16 and 2.17 show the result of passing the filtered 90* and 180' component seismograms through a response spectrum algorithm (5 percent damp-ing). The dashed trace gives the response spectrum of the input seismogram. The heavy solid traces show the result produced by the zero phase filtered seismograms. The other two traces give the results for the observed and random phase seismograms. The zero phase spectrum almost completely envelopes the other two. In each case the zero phase spectrum is highest and the random phase spectrum is lowest. These figures depict the sensitivity of computed response spectra to the phase shift assumption. I i 2.3

1 1 300 200 100 '.^yv". ^ ^ r -- 1--

0. O i;

y -100 -200 -300 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 86.E01.1979 INPUT SEISMOGRAM RUN NO S27 PLOT NO 1 I

l l Figure 2.1 90-degree ccx:ponent of the October 16, 1979 earthquake recorded on the USGS accelerograph pad on the dam abutment. l l [

3. 0

2. 5

2. 0

1. 5

1. 0 i

0. S i

1 1 1 1 I

• b. 0 50 100 150 200 250 INPUT XFER FCN RUN NO 527 PLOT NO 2 Figure 2.2 Zero phase transfer function, Auxiliary Building foundation /accelerograph pad, transverse cmponent,

'Itst 4, shot 4. Zero phase is plotted as the line at a value of 1.5, while the other line is the nod-ulus of the transfer function in units given along the ordinate scale. Abscissa scale is in Hz. c f

l 400 300 200 100

0. 0 1

I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO 527 PLOT NO 3 Figure 2.3 Irgulse response of the zero-phase transfer function shown in Figure 2.2 (Auxiliary Build-ing/ pad, transverse ccuponent, Test 4, Shot 4).

Abscissa scale is in seconds.

80 60 40 O. O ^ "vY p"y,y//.V,'i t ^ = ^ ". -20 l -40 -60 -80 1 I i 1 1

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 527 PLOT NO 4 i

l Figure 2.4 Auxiliary Building foundation accelerogram obtained by filtering the 90-degree ccrnponent of the 1979 l accelerogram using the zero phase trans"er function shown in Figure 2.2, Abscissa scals is in seconds, i l

I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO 529 PLOT NO 3 Figure 2.5 Impulse response of a randcm-phase transfer function with the same spectral nodulus as in Figure 2.2.

Abscissa scale is in seconds.

30 20 ~ \\ 'g P.; [hM.'FF. #.

0. 0 c

-1C j l -20 -30 1 I I I I

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 529 PLOT NO 4 Figure 2.6 Auxiliary Building foundation accelerogram obtained by filtering the 90-degree camponent of the 1979 accelerogram with the randem-phase transfer function whose inpulse response is shown in Figure 2.5.

Abscissa scale is in seconds.

400 330 200 100 ,"j !..

0. 0 t I

I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO 528 PLOT N0 3 Figure 2.7 Impulse response of the observed-phase transfer function, Auxiliary Building foundation / pad, transverse ccrnponent, hst 4, shot 4.

Abscissa scale is in seconds. 1 l ,n-e -n,

60 40 20 S f t".h-Y, ^7,h y yryj,VM yi.'.-.T -- ^^ O. O i -20 -40 I I I i 1

0. 0

1. 0

2. 0

3. 0

4. O FILTERED SEISMOGRAM RUN NO 528 PLOT N0 4 Figure 2.8 Auxiliary Building foundation accelerogram obtained by filtering the 90-degree ccrnponent of the 1979 accelerogram with the observed phase transfer function whose 2mpulse response is shown in Figure 2.7.

Abscissa scale is in seconds.

300 200 100 O. O I ' 'S A ^- %"^ ^ y7-v -100 -200 I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 B6.E01.1979 INPUT SEISMOGRAM RUN NO 512 PLOT NO 1 Figure 2.9 180-degree canponent of the October 16, 1979 earthquake recorded on the USGS accelerograph pad on the dam abutanent.

I-

400 300 200 100

0. 0 1

I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO 512 PLOT N0 3 Figure 2.10 Impulse response of the zero-phase transfer function Auxiliary Building foundation / pad, radial emponent, Test 4, shot 4.

Abscissa scale is in seconds.

100 50 "n~' 'q r p"' i rv"i - - - r" ^ ^ ' - '-~ " ^ * '"

0. 0

-50 -100 I I I I I

0. 0

1. 0

2. 0

3. 0

4. &

FILTERED SEISMOGBRM RUN NO 512 PLOT NO 4 Figure 2.11 Auxiliary Building foundation accelerogram obtained by filtering the 180-degree cmponent of the 1979 accelerogram with the zero-phase transfer function whose inpulse response is shcun in Figure 2.10. Abscissa scale is in seconds.

I I I I i

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO 514 PLOT NO 3 Figure 2.12 Impulse response of a randcm-phase phase transfer function, Auxiliary Building foundation / pad, radial cmponent, Test 4, shot 4.

Abscissa scale is in seconds. i i l l

40 "( kIl'l - l I' ",'1'$^ d di!#1 ^^- 8 ; - 'o 00 -20 -40 1 I I I I

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 514 PLOT NO 4 Figure 2.13 Auxiliary Building foundation accelerogram obtained by filtering the 180-degree ccuponent of the 1979 accelerogram with the randcm-phase transfer function whose impulse response is shom in Figure l

2.12. Abscissa scale is in seconds. l l l i w

400 300 200 100

0. 0

.$ ' '/,'. I I I I I

0. 0

0. 5

1. 0

1. 5

2. 0 IMPULSE RESPONSE RUN NO S13 PLOT N0 3 1

Figure 2.14 Impulse response of the observed-phase transfer function, Auxiliary Building foundation / pad, radial ccruponcnt, Test 4, shot 4. Abscissa scale is in seconds. i i l i l t m

o 50 AAA ' ' " " - - " ^^"''""#d* ^ u 'p g y vi"I ~ ' ' '" ' ' " ' ~ ' ' 'i"' ' II Ilv g lp' y "' ' ^ - O. O rr -50 1 I I I I

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISM 0 GRAM RUN NO S13 PLOT N0 4 Figure 2.15 Auxiliary Building foundation accelerogram obtained by filtering the 180-degree ccuponent of the 1979 accelerogram with the ob-served-phase transfer function whose inpulse response is shcun in Figure 2.14.

Abscissa scale is in seconds.

O O O O l e o s s s s o-3 A / x' MN , j[y'QMpfg s/'x3jjfE,'Afx['i'jjM'{p,'Xj O y s vuxwu / v x vuxww / v x vwxwv - s NX /1\\ XfNYXMbX / \\- XAXYYXNnhX / \\ XXh XXMhX / NNX I W/WXXhKNSX '>WMXYWNNX W/MXXhKNNX X NN h W /J WX!XMMhM'4WXLX'Mshv'#M/X{WAN\\% y NMh)< >hNbN/)( >h\\ Nh)(>INM O S / , us .s y C / <T / \\ ft \\"; T Pv'XW W / /\\ / \\ +1 N6\\ N'y </ J ' / /\\ / \\ /N \\i \\ Vt'< v' I i/ /\\ / / A A \\/ Mh'NY&\\!A / A A \\ b &NNVt/ /A A \\ \\ N'(VM / A \\ w ~ /V \\ / WV'XMWhi / v \\ VNkXNW / V \\ V/VV X3W /\\ f E sX / \\ iWXKOKhM / \\ XXXIX'XMbX / \\ l X %tX'MhX / o NNX W/NTxN>M X W/X@%KNNX W/hXXhKNNX z %%SIO!XDU%SV)(!XXAk%l/IP)(!XN& NM'[OINMY)OINM'/000xW 'o ~ $9 b[)sM b!!MinCad /,Xg \\fy%MPjXg i'fd'j;$pp,'X ff y'j;jQ','Xf o ~ f l r-yal v wemu / v x vmww / v x vwxwv / 3 W W //A Y /'%%^XhW / \\ X/XfXYNAKX / \\ MfXrXX4XbX / NNX //IV W/hMWX3X WXGXXsKNNX W/hXXsxNNX o @'/4NI#XW)4'i<\\MNN#/PX/X;'RNh%/MWXX(%% 8 NM3)<}INE%3^zOCxW6)( '>IN$$@ \\ / E-M 9xCr Ev Xs' 79xC sM @'ti'/!'LMSEl'7/;'/N'330l; '// '/!'IU$N623'7/ v x vwxwv /v x vexwv/ v s vwww/ 3 K / \\ X/A/X XhshX / \\ X/NX XWhX /l\\ - XA.XfXTnsX / NNX wzOxXs>5xX WZOX XN>NNX XvhXX%\\X X\\%/MW>4X'.M%#/fWXLM1XN%/4WXXKh% O NM))OiNMY)< >(NMY)OINM ' / 9 M lxCsM 'CvM C#M 0 0.01 0.1

1. 0 10.O PERICO IN SECONOS RUN 467 1979 90 COMPONENT Figure 2.16 Auxiliary Building foundation response spectra obtained frcm transverse ccrnponent transfer functions for 'Ibst 4, shot 4 with randm phase, observed phase and zero phase (in ascending order). 'Ibp trace is the response spectrum of the unfiltered 90 cmponent 1979 accelerogram.

O O O O o o s. s g s s o. MK 3' '""O MJ 9o .s,:s x NN's'u,, :/ x, N x 'cs'; :,, w's, N x NN'su. ;: x - ^'^'N "4 "'a ^'^N+'"*' ^' ^ ' ' N +"+ M ' ^ a ' / Y \\ N ' N'M N NW / Y \\ N'hM XNW / Y YNM VNW /\\ sX /i\\ l V/^f X '<T, ANV /1 \\ l X4A(X X'%bX /f\\ L,V/A/X X4AbX /

sv i wx4>rxmNx I x+m xbxxy wx5x xkxxsx XN%V//FXl/M hYs\\% 6///X/M DXN%#dWX{ X'fM\\%

d / NMY)MKDMy7MDMY)(X's3 O O M9 / x ,s N x 'dev/W;: x. N X' Jc0 :w; 'x,AAX '.c c 0 : 4 Y,' x - n o A A is, wawe < es a,sx+wsw - A s n +ww- - A '/Y N /4'/(AXNW / V N YNXX4W / V N N'h X XYtN/ / ' E NX /\\ lMMXYWX / '\\ 6 Xa^<XXMW / \\ X4NX 'OAsX / o xw wwxyg3xx x e<gxxmsy >ww xx%x z M N's#d W>{XNMx%#/WX XNXN%#'WX X%% sW3D< ')(NM'))ONM6)ON $$@ 'o ~ $9 kMk MM8 shMbsb /;//N'ur40c m 'x.x';'cs"0 w, ^x, x x'c ' MW ^< / o o /DilA W ANWWeiW A t A 4 x NNxVW/ /A A \\ NNXVVs/ / ^ 'I / //fl W W /f /\\ . N'W'M XNVM/ V N N/NXXVW / N' \\ Y/t'N W Y / \\ MAfXXWsH / \\ I XrXfXXWsX / \\ MAIXX1XbX / y

ss'/s/ I WMWdh W Xfx XhKNNx WX1XXKM'<

a >@/sb'%#4X/X'XNX'\\h#A W)4 MXN%#2W>4X%% @M6)<X s E>6)('>('NMY)<['>N$$8 / E-N Rb ll[' lsbsk o x, s x'Js';m WX A X'Ja:r;' 'x A X'A'0+'E:/x - - A ^ i s s umu - ^ ^ i s sawWa e ^, ^ e x N +xwe - ^ ( / Y \\ NWXX4W /Y N VNXMMW / Y \\ N'NM MNW / N sX /\\ X/AfX X%hbX / \\ lX4XIXWnbX /l\\ lX4XTMY %hX /

NW MX XXbKNNX WXfXXsxNNX 1 WX/X Xh&\\Nx XN%//A'X/MM'fM%#//X/X'k'N SN\\V/d M/MNA%% o b

/ sx' M,'D< )C'NM69O('NMY'MfsM '/ sg /N l >'/ /'xj y / N j . O 9 hk k. ' g' b d N d N f d,c 9 d 0 0.01 0.1

1. 0 10.0 PERIDO IN SECONOS RUN 461 1979 180 COMPONENT Figure 2.17 Auxiliarf Building foundation response spectra obtained frmi radial cmponent transfer functions for Test 4, shot 4 with randcrn phase, observed phase,and zero phase (in ascending order). Top trace is the response sp%h71 of the unfiltered 180 cmponent 1919 accelerogram.

Question 2A. Regarding the sensitivity of the results to assumptions about (New Q. 2) phase response, the Licensee should calculate upper-bound reduced response spectra for computed foundation motions that have been processed with a spiking filter (e.g. Robinson, 1978); the processed signals will have the same emplitude spectra but will be compressed in time as much as possible. Alternatively, the Licensee should demonstrate some other upper bound, given that the pad-to-foundation transfer function may not represent a physically realizable filter.

Robinson, E.A., Multichannel Time Series Analysis with Digital Computer Programs, Revised Edition, Holden Day, San Francisco, 1978.

In assessing assumptions about phase response and the relative durations of input and output signals, it is relevant to compare the durations of founda-tion and free-field signals. Such a comparison is presented below, and it is found that there is no systematic difference between foundation and free-field signal durations for frequencies above 15 Hz. Below about 15 Hz the free-field durations are longer in some cases, but this is unimpvrtant because the reduced foundation response spectra lie below the M 4.5 RIS spectrum at g these lower frequencies. The observed similarity of the signal durations justifies the use of the zero phase shift filter, which preserves signal duration, rather than any other filter which compresses the signal duration. The premise of the question is that, as an upper bound case, foundation signals should be spike-like and compressed in time as much as possible. How l this case could arise physically is not apparent. Both at foundation and free-field sites, the seismograms are composed of P and S groups travelling at bedrock velocities, followed by higher mode surface wave trains. Multipathing and reverbration due to the heterogeneities along the propagation path and i near-receiver structure consplicate the signals and extend their duration. The character of seismic signatures is alike for the deeper and shallower RIS i l l 2A.1

a events, and for earthquake and explosion sources, as shown in the response to Question 1. Because the seismic signal at any receiver location consists of a multiplicity of arrivals at intervals determined by the velocity structure, it is not reasonable that a spike-like seismic signal of arbitrarily brief duration could be incident upon the structure foundations. Such spike-like signals f abricated with special* filters do not provide physically reasonable l bounds for the foundation response. The Robinson spiking filter is inappropriate because it does not preserve the amplitude spectrum. A brief description of this spiking filter and reae examples are given below. An siternative filter that does preserve the amplitude spectrum, and that gives the largest possible peak amplitude, was used. The phase of the output i signal is set to zero at all frequencies. This produces a symmetric, non-causal, band-limited impulse that is the most spike-like (maximum amplitude) signal that preserves the original spectral amplitudes. This filter, called here the impulse filter, produces signals that have response spectra only slightly higher than response spectra for the zero phase shift filter, for frequencies below 40 Hz. However, the zero phase shif t filter is preferable ~because it preserves the character of the signal, whereas the impulse filter is physically unrealistic; it requires the output signal to be unrelated in phase to the input, spike-like and symmetric in time. Comparison of foundation and free-field signal durations The L-22 velocity re' cords of the explosion tests generally have longer dura-i tions at free-field sites than at equidistant foundation sites, but this does not mean that the durations are longer at all frequencies. The band pass filter results presented below show that for frequencies above 15 Hz, free-field and foundation signal durations are nearly the same. For frequencies i below ' 15 Hz, the longer durations of the L-22 velocity records at the free-field sites are indicative of lengthier higher-mode surface wave trains: note that the velocity spectra of the explosion records have peaks in the band 10 - 15 Hz so that these lower frequency contributions control the duration of the observed broad-band signals. 2A.2

Figures 2A.1 10 show Hilbert transform envelopes of band pass filtered free-field and foundation seismograms. The band pass filters have width + 6 Hz to the -30 db point, and were computed for center frequencies f rom 10 to 32 Hz at 2-Hz intervals. The numbers to the right of the traces give the maximum amplitudes of the envelopes. The envelopes for foundation and free-field records are scaled to the same maximum amplitude in the plots. Auxiliary Building durations for Test 3, shot 2 are compared with those for free-field sites F1, FR, F6, and F3 in Figures 2A.1 - 4. For frequencies above 15 Hz, there is no systematic diff erence between Auxiliary Binilding and free-field durations. The same is the case for the Diesel Cenerator Building and Service Water Pumphouse data, shown in comparison with free-field site F3 in Figures 2A.5 and 2A.6. Note that in some casca foundation and free-field durations are the same over the entire band 10 - 32 Ez (e.g., Figure 2A.4, transverse component, and Figures 2A.5 and 6, radial components). Band-pass-filter envelopes for Test 4, shown in Figures 2A.7 - 10, indicate that foundation, free-field, and USGS ' dam abutment accelerograph pad signal durations are essentially the same above 15 Hz. and in some cases for all frequencies in the band 10 - 32 Hz. The similarity of foundation, free-field, and accelerograph pad. signal dura-tions justifies the use of the zero-phase shif t filter, which has no effect on signal duration. While in some cases the free-field durations are longer for frequencies below 15 Hz, this is of no concern becsuse in ali cases the reduced foundation response spectra fall below the M L 4.5 RlS s pectruth at these lower frequencies. The Robinson Spiking filter Robinson (1978) has described (and given computational algorithms for finding) a " spiking filter," which ideally produces an output signal from an input signal in such a way that the output is a spike located at the most energetic part of the input signal. The spiking filter is intended to resolve the arrival time of a wavelet, and does not preserve the amplitude spectrum. The 2A.3 _n-,

4 output.of the filter approaches a value of I for the ideal case of a delta function. The spiking algorithm ignores spectral amplitudes, so that, what-ever the input, an output near unity for the spike is obtained. In the examples given below, the spiking filter output is normalized to contain the same energy as the input seismogram, so that the effect of using such a filter can be investigsted. In Figures 2A.11, 13, and 15 are shown the vertical, 180- and 90-degree components of the 1979 accelerogram as recorded at the dam abutment. Given in Figures 2A.12, 14, and 16 are the spiking filter outputs obtained using Robinson's algorithm. Note that 256 datum points were used in the construction of the spiking filters, and that the height of the resulting spike is a sensitive function of the number of datum points used in constructing the spiking filter (in this case 20 samples).~ Therefore, since the resulting high-frequency response spectral output scales with the peak amplitude of the input seismogram, the hign-frequency asymptote of any response computed will depend critically on the length of the spikirg filter constructed. In any case, this type of spiking filter is totally inappropriate, because it does not preserve the signal spectrum. 4 Impulse filter In general no single selection of phase will give a response spectrum which is ' highest at iall frequencies. The phase ar.gle selected here is the one which gives the largest possible peak ground acceleration on the output filtered seismogram, namely zero phase at all frequencies. This filter, called here the impulse filter, 'is constructed by multiplying the Fourier spectrum of the ~ input seismograms by the real-v. slued spectral modulus ratio, then replacing the.' resulting complex-valued spectrum with real values consisting of the Fourier spectral amplitude. It is clear that, for a signal with fixed Fourier amplitude, the impulse filter gives the largest possible peak amplitude, because at zero time, the amplitude consists of a real sum of all of the Fourier components, adding together in phase. 2A.4 T 4 -- -- m i, %. ,e_. ' - - - - - - - - - - ^ ^ ' - ' - - ' " ~

In Figures 2A.17 and 2A.18 the impulse filter results are compared with those for the "zero phase shif t filter" for the 90- and 180-degree components of the 1979 Honticello Reservoir earthquake. Unlike the impulse filter described in the previous paragraph, the zero phase shift filter sets the phase of the filter, not the output seismogram, to zero. Thus, the result (top trace in these figures) retains the phase information of the output seismogram. In contrast, the impulse filter seismogram has a Fourier transform with zero phase at all frequencies, and therefore the seismogram is symmetric about the origin, where the peak amplitude occurs. Notice that this peak value of i acceleration for the bottom traces is almost twice as large as for the zero-phase shif t ' filter results (top traces). Figures 2A.19 ar.d 2A.20 compare the results of passing the filtered seismograms through a 5 percent response spectrum algorithm, for the 90- and 180-degree components of motion respec-tively. Notice that the impulse filter response is higher at the very high frequencies, but that gene rally the two traces are fairly close together. From these calculations it is evident that the choice of zero phase shift or impulse filter does not make a great difference in the computed response spectrum below 40 Hz, although the latter lies generally abcve the former. Further comparisons of results for the zero phase shif t and impulse filters are shown in Figures 2A.21-23. Foundation envelope response spectra are shown for signals obtained for transfer functions constructed from 50th percentile spectral modulus ratios for the Auxiliary Building, Diesel Generator Building and Service Water Fumphouse. While the zero phase shift filter preserves the character of the input sig-nals, the impulse filter does not, requiring instead that circumstances conspire such that the output signal has zero phase at all frequencies. This is not physically reasonable and gives an unrealistically conservative esti-mate of the foundation response. e s, i 2A.5 i-v s c. r

10.O No 102. COMP [.,,m 10.O R No 106. CSMP[. ,/ \\~s a.sene / \\

a..=n

_ _ = ~ 12.O 12.0 f ,rl \\ / sg e.assee / \\ a.esste / xn s 14 0 14.0 onn /\\ / N U \\ a.comes ~ _/ \\ a.seest 18.0 16.0 l \\ I t / (, a eases g/ e, e.earse 1 s. 0 l a. 0 \\ a.seems a.estee 20.O 20.O, \\

s. N -- --.

E"'."s s 22.0 22.O 7 f s \\ -s-__""*. ~- 24.O, 24.o f \\ // \\, a.neep

e. seats

,_ __s as, g 2s. g I\\l \\ e. n \\

s. t**)

i 2s. 2s. 0 g / \\ \\ _ - %w }'& j 30.0 30.0

s. easts Aeossa-v 32.0 32.0

\\ /I e.neen I N_ es% see n, ens arte na ens annt as ens arte at ens annt D. 00 0'.85 1171 2'.56 0.00 0'.85 1.71 2'56 n GURE 2A.1 Envelopes of band pass filtered records for M 3, shot 2, at fne-field site n (solid lines) ard Auxiliary Building (dashed lines). Inft, radial; right, tra.m.

10.0 RUNE 102. COMP 2. 10.0 No 106. COMP 3. L ses3e T a.eas44 I ^ \\ / L Ear" f 12.0 12.0 7 jS

s. massa a.seen

\\ \\- / 14.0 14.0 L e3ess O E ess22 \\ l \\ a.easse a.esset v ~ 16.0 16.0 a.eassa a.sesso ( s t a.eam 18.0 18.O e.signe /, a.sessi I

a. eases s sense 20.0 20.O a.soses a.easso

\\ a.sene s.co:47 ( s 22 0' 22.0 / L es327 a.se2M L selet

Ly4, 24 3 24.

I ^ O f__~_ us_ as. 2s. 3 1 P A __ b**? 0 l 28. 28.O g a.estse a.setet \\ ,,/ 2 *"' ~ 1 A _ _ 3-M 30.0 30.O a.ee m r \\

a. senn

\\ / n~ _r x_A

32. 0 32.O

\\ I IN \\l ' " = ' n_ >== at ens nrat at ens mant at eno arnt at eno annt 0.00 0'85 1.71 2'56 0.00 0' 85 1.71 2.56 i l FIGURE 2A.2 Envelcpes of bard pass filtered records for hst 3, stet 2, at free-field site FR (solid lines) and AcKiliary Building i (dashed lines). Inft, radial; right, tem.

I 10.O UNe 102. COMP {., tota 10.O MUNe 106. COMP 3. j a,esoe ,/ I \\a~ = s.eam n

12. O

12. O "N j

a.easet s.eeses v /- / ' a.essee \\. s \\ ',,,, ~ ~ m s esete / ~ m 14.0 14.0 f a, ens 1 a.neest j\\ 'N a.seese c/ vm_ s.ensit ~~ 16.0 16.0 g !( b 18.0 1s. 0 \\ / a.eMas 3.es1es w I n 20.0 20.0 \\ a.sorte - - s, a.ca 1

22. 0 22.0

\\ ,/ s

e. eu.i m -

-__ none. 24.0 24.0 \\ l a.eens yy s. 7 c.eeejs. 2k o 25-e.,e e.m \\ a.comes \\

• \\

a.eens? J vu _-_1x g

==; o

==. o s. a.ees,. I l \\ j o e -f_s _ _.1W s 30.0 30.O r \\ \\ c ~ __. x,.ht*** 32.0 32.0 l / N. an ens arnt at ene mont at ene arnt at ene m g 0.00 0'.85 1'.71 2'56 0.00 0.85 1.71 2.56 FIGURE 2A.3 Envelepes of band pass filtered records for hst 3, s.ct 2, at free-field site Ff foolid lines)and Atzxiliary PdWM (dashed lines). IA*t, radial right, transverse.

10.O RU 102. COMP [.,,,, 10.O No 106. COMP 3. j\\ j a.aones lI %M _ '2*** ,I ^, 12.O,\\ 12.O p f (%_ / = f 14.0 14.O 7 p I I \\

a. e==

\\ n. en v s., m 16.O g,,, 16.0 \\ \\ ^m-18.0 18.0 / m _ j1 \\ ^ 20.0 20.O / \\.

a. cons a.es 47

( 22 o 22.0 r a.aone p u I *2* ~ 24 0 24. \\ \\ \\ // a.sens pf a. or 26.0 26. I i /\\f ) \\ \\ y _ m U,, i 28.O 28.0 l / \\ E s 30.0 30.O r \\ I "W A - r a _^ _ xW. '.

32. 0 32.0

\\ I 'Y / hO at tno arte at ens maat at ens.nrte na ens annt O.00 0'.85 1.71 2' SS 0.00 0.85 1.71 2.56 i'IGURE: 2A.4 Envelopes of band pass filtand M M M 3 M 2' at free-field site F3 (solid lines) and Auxiliary Bdiding (dashed lines). Inft, radial; right, trans"erse. a-

o 10.O U 102. COMP {. 10.O No 106. COMP s. I r I V / i/ m _~_ s.neene

a. o e

12. D 12.O A

c. / ./ _~ v s.sessa \\ e s _ _ a.e:ese m 14.0 14.0 \\ F./ O / \\s a.seest a.neen 16.0 16.O V< b _ '.*.*"4 \\ s 18.0 18.0 f / \\ Il\\ / \\

a. e==

,/ e cayes 1 20.0 20.0 3 "s'eaut4 s.so44s 22 O 22.0 n / jV N ' '*.h

a. "*52 24 0 24.

3 r \\ l V

a. D s

~ 2s. 0 26. J lm / /\\ g cs / q,,,, 7_ 28. 28.fi / \\ \\ / / V /[c& ,/ JJ.'. 80 / fg

a. = ni

e. e==

80 O ( %i \\ ~ I. ,/ w x ngunst s 32.0 s2. 0 \\ r' 1 \\^ /s A // \\ uw

== s-e na ens arte at ens ante at sne.nrte at ens an== 0.00 0'.85 1' 71 2'56 0.00 0'85 1.71 2'56 TICUIE 2A.5 Envelopes of bard pass filtered records for hst 3, shot 2, at free-field site r3 (solid lines) and Diesel Generator Building (dashed lines). Inft, radial; right, transverse.

10.O RU 102. COMPg. 10.O Ne 106. COMP 3. /\\ / \\ /s h, ! Q,,- \\n _, 12.0 12.0 T 8.Wese g A sesas O / /\\

t. eNoI

} sense n 14.0 14.0 S.ettee S

8. sects

/ \\ \\/ \\ / 0.env30 V R.ges15 s x x-16.0 16.0 g a.e saa a.setes g \\ \\ s.sossa a.coane >s N__ v m 18.0 18.0 a.sene i n.sesse g \\ ~/ _ _ = >.. so m s -_i s u 20.0 20.0 g 7\\ N \\

s. j

,u/

seie, sees, 22 0 22.0 3

/ -s_ 3 d 24 0 24.

s. seine

'\\ a.ootse gi 7 / / \\ ^

• k 5

Y MN

1. ^

v so 26. 26. N. s.osan A s.soaei f \\ \\ \\ U '. T J 2w n'_* **l_ J ~' 2s. O 2e. 0 S. M304 M 8.se1e7 /\\ \\ s-J 2_^ % AR ? 30.0 30.O a.setta g a.seano / N s. . _ _ m ""4 32.0 32.O

e. coast a.aenes g

seser

e. sects g

g .J w .a at ens arte at ans ampt at ene armt at une awet 0.00 0'85 1'71 2'.56 0.00 0'.85 1'71 2'56 l TIGURE: 2A.6 Envelopes of band pass filtared records for hst 3, det 2, at free-field site F3 (solid lines) and Service mtar l P.msticuse (dashed lines). Inft, radials right, transverse. [ l

i io. o % nun. io4. cone g.,u 1o. o pun. ios. cone s. f i J ss s s, 12.o 12.o /\\ b n sts ~ \\s if

1. S*d*

_/ > ~ s 1'A0 1' O g = f l\\ \\ ^ / \\ </ / / ""3 N' 5 v 4 'J'!'". 1 o 16.o s lg /\\ s.: / V \\/ %._sanon. \\ - % us s I a. o la o, f q v / N _ ~ ss*' ""' 20.o 20.o, 7 h'g / \\ g E 22.o 22.o /g \\ / w r M ? *2% 24.o, 24.o ,,,,n / rI \\ <~ f - :n *~ Ni

• ' **Z 26.o as. o,

I = / \\. ~ r c._ a 2a. o 2a. \\ \\ / v .=2. w_ -s 80 0 so. s/ \\ /\\ 3 as . y. 32.o s2. o / Il/S/\\ AN / u na sn= ap>< na ent mq me ens aps,e me ens nn.,t

o. co

c. 85 1.71 2.56

0. 0o

c. 85 1.71 2.56 I

FIGURE 2A.7' Envelopes of b.nd pass filtared records for M 4, shot 3, at the IISCS wr=1-%4 pad (.olid lines) and 1"v414a7 P'4 WM (da.had lines). Inft, radials right, traraverse. l l

f 10.O UMe 104. COMP {. 10.O UN= 105. COMP 3. / r r .~ ~., 12.0 12.0 / / NM / \\A s 84 0 1' O 4 fg .== f I' ) "a ,M >=g > = 0, l l \\;~ p s anc*'. v

a. eeus 38 O 18 O r

a. m=>

r mim 3 N / l \\s /g-V \\^ S "2 20.0 20.0 sorse f a.eotr2 / m \\ i e4 M 4 c / - ' ""1 r 22.0 22.0 \\ -A**. ~/ m*'9'* s 24.0 24.0 / \\ c. ^^ M""*, N 2 \\ s 2.js 2.0, J \\,v u t u s> 28.0 28.0 /v/ /\\ soin seaso g / ' ~

• v u

~ :. e s s 80 O 80 O f .uu fg min / e1 \\e 2._"- 32.0 32.O g I \\ S Nr d "e*" ( m na snt arts na. tnt nant na snt anc na ent anne 0.00 0.'85 1.71 2'56 0.00 0'85 1' 71 2' 5. TICURE 2A.B Dwel: pes of bard pass filtared recx3rds for 'hst 4, stot 3, at free-field site FS (solid lines) ard Auxiliary Buildirs (dashed lines). Im*t, radials right, transwese.

10.O RUNE 104. COMP {. 10.O R(No ICS. COMP 3.,,, p [ 1I ^ \\ ^ N==a y y/ tr m ^ sc y 12.O 12.0 ,3 5 3 ,,,,m j {/ -haaaae J / a. y y 14.0 14.O 7 E _[ m J v 16.O 16.O j\\ Amus a / -[ v \\/\\. /k D /\\ve g 18.0 18.0 p J\\ r\\ M "'" (S/ /

• M c

n 20.O 20.0 f \\ / / %nl \\ A mg ( f

e. a 22.0,

22.0 r T h /^ a.coYo 24.0 24.0 7 ,,,,, n I \\ r \\r '" 1 V l \\ W ^ i m t 26.0 26.0 g g,,, I f \\ ./

a. cones

\\ - a.g m / 28.O 28.0 \\ 7 ,,,,, n \\s e.90V

a. gile s

y 30.O g,,,, 30.O, J ,j Ay \\j ./ gj 32.0 32.0 g,,,,, I ,e I L u^l N 74 (f I \\rw"\\ e M. n m na ens negt a4 =n* mnts na ent anss enn a n,s n1 0.00 0.85 1.71 2.56 0.00 0.85 1.71 2.56 FIGURE 2A.9 Erivelcpes of band pass filtered recads for ' nest 4, shot 3, at the IEGS accelum.,6 pad (solid lines) ard Diesel Ge:neratce Building (dashed lines). Is.ft, rahal; right, transverse.

10.O UNe 104. COMP 2. 10.O M e 105. COMP 3. k oesn

a. eases in 3

/ 5 \\. ( 12.0 12.O a.eua4 T a.sens f g f _, J l l \\ S.)e%ef 14.0 14.0

a. eses *

/ e seaso e n / g M h e.g v m 16.0 1s. 0 a.nesia a.mossa \\ \\. A \\ f \\ -?;j W y t s e. */**_ 4 18.0 18.O J / ) \\*"*. C~ 8 ~ 20 O 20.0 r i. / /*~ I ^ ~ g " -- N "Amely {N e

a. e

22. 0 22.0 7

g /\\ /* U *M / \\- 'P / m ~ 24.O 24.0 f \\ r \\r ~. A'*-- ?J \\ \\ f.:t% 26.0 26.O q /\\ ~ ^--._ ~ Le l 2e. 0, 2n. 7 \\ \\.n___n n O '** ?l J 'Y 30.0 30. p ~~ J \\, A. y ,1 J~ ag. ~ l l 32.0 32.0 7\\ \\ \\ \\/f f j \\, v ' y " L%esee, [ na.eni nrgs na enn ante na enn ness na enn anec 0.00 0'.85 1'71 2'56 0.00 0'85 1.71 2.56 TIGURE: 2A.10 D:velopes of band pass filtered recxrds for hst 4, sh.t 3, at free-field site F5 (solid lines) and Diesel Ga.cew Building (dashed lines). Inft, radial; right, transverse. t

17551.GGGGGGG A \\ h 1 I I i I m v v f V h i -17351.GGGG000 FIGURE 2A.ll vertical c2nponent of the 10/16/1979 tbnticello RIS event recorded on the USGS accelerograph pad at the dam abuta nt. (first 0.512 seconds)

34855.5015525 1 i J\\ ) n =~v y V YUNV r Ij l I f I I ( i -17225.8593750 FIGURE 2A.12 Spiking filter outp4t for vertical cmponent of 1979 Monticello event. Spectral anplitudes scaled to provide same total energy as Figure 2A.ll.

mm a 5 f 345S5.0000000 f O _ /\\ t v VV / y. I b) J 4 1 -23734.0000000 FIGURE 2A.13 180 cx2nponent of the 10/16/1979 Monticello RIS event (first 0.512 secondsl, f ~ - - g

53507.S295375 l 1 .I.I i 't i I:i 'i l 9 s O f f n n \\n b-Mi e \\. l a, -_-W b y i v v ] \\ J t il 1 i 1 -24174.1875000 1 l l FIGJFI 2A.14 Spiking filter output for 180 vui p ent of 1979 Manticello event. Spectral a plitudes scaled to provide same total energy as Figure 2A.13.

4 34575.C000000 h l ,O 8 ^v h n J -34375.0000000 FIGURE 2A.15 90 W.e Eent of the 10/16/1979 Monticello RIS event (first 0.512 seconds).

59395.2573125 i I l i I l n lI f b f ^ W^v y j g 7J e,7 v.A 7 l } } i j h i f -22119.2148438 \\ FIGURE 2A.16 Spiking filter output for 90 omponent of 1979 Monticello event. Spectral auplitudes scaled to provide same total energy as Figure 2A.15.

100 50 , ',] .j., lf,^{. ^,-,:l.

0. 0 L^

4 I -50 -100 1 I I 1 1

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 804 PLOT NO 4 200 100 i

0. 0

...aAIdb $ kEA$ hl 4 t.m.. -,, cy pg - g g,,., -100 1 I I I f

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 805 PLOT NO 4 Figure 2A.17 Cmparison of filtered 1979 RIS Monticello event, 90 a:rnpanent, using zero-phase (above) and impulse filter (below). Both used AB/P2 zero-phase transfer function.

1

a 100 30 d 11-O. O ""y g pyo. -50 -100 I f I I 1

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 802 PLOT NO 4 200 100

0. 0

lllhy.

,Ylf'l,. -100 1 1 1 1 1

0. 0

1. 0

2. 0

3. 0

4. 0 FILTERED SEISMOGRAM RUN NO 803 PLOT NO.4 0

Figure 2A.18 cm parison of filtered 1979 RIS Monticello event, 180 cmponent, using zero-phase (above) and inpulse filter (below). Both used AB/P2 zero-phase transfer function.

D D N N O N s o- ~ N g o L s,nx'A's',/x,n xvs't+ %',L'x,a, a' m N n ^x / 1 ^.A x wmu - a s, a\\ s sw<w a s rs w < w - A V Y \\ VNXYWV / V VNX XNNY / V x VNXYnY /h W /I\\ X/NXWsW / \\ MMX XMW / \\ XfNX XWhX / j hv i Xa ys4x D KNNX xzu>rxhv VMxD>w XhN##M/>O<NA%##A/>OGA4##//x)<NN f N2OOM3^OC M 30<XW o o e Nhbf AliCs!Nthbs! N o I'?/l'l 'b $ U'?i: 'i '!NN'?/ ']N$$$II'?/,' D N /- v x vuxxwv - v x v~nvu - v x vexwv - E AX / \\ X4NX M AbX / \\ MM

X'WNX / \\

X4X>'XMbX / o sxx pfxygmx ymqyx3pm pfpyx%y z A%/43/xXA%#2%/xD<NA%/4%/xMM NMEC<l>INMYCINM6)INM 'O ~ "+ s e M9aCsM8tnC:!R$$nC:!M ',7jly','jggy^x,Nj',jgg'/x, \\);xjyy^A, o A / / /V x VNYXwT / V N VN>SNX mV / V x vN)'~>MV / 3 d .NX // \\ X/ M:N / \\ MM3MsX /8\\ %AO3MsX / hVi W/XbM >M9XX5?w XM9XX%\\x o m ee wxwAwaw/x <NA%Aw/x' '<N>tm S NMY9(S>INNMYCCNMY)CNM / E - Nhs!bhM89EC :MR$kOM* 77jN'4','ggg'/Xjs):3gg g'fxj N,X;ygggg,'Xj a r-v x vuxxwv - v x vux>cwv - v x vunwy y NX / \\ IX4NXXVQsX / \\ X4NXYWhX / \\ X/V X' N / hv xzsixx4v w/NXX9m >E9xxmNX % ##A/>C<Nh%##A/x%%%f#//xMh% o N >?)INM$69INMEC<S)(NM ' ' Ebkhb!sbhthhbl!bhhhbi 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS RUN 902 Figure 2A.19 Cmpariscn of response spectra for foundaticn signals obtamed from the impulse filter (solig line) and the zero-phase shift filter (dathed line). 'Ibe 90 wwent of the October 16, 1979 accelerogram was filtered with the transverse Auxiliary Building /

accelerograph pad transfer function for Test 4, shot 4.

7 D N N V N s 0-t R k ?$$ 8 '?/?'2R'l#Si3E 'X/ ','i\\dN R b'X RR'83}M;P'X ~ vme<w /, r- + x vowwy - v s ve my v s sX /f\\ 1X/XfY M \\Y / '\\ X4X/X TM X4h 7'YWNX / w w/MxwKw wmxww/\\wwxxww M

1. A%/>C<M4##</xM>t% ##' </x 3<0KM f

_N MY9<[NNMYCD('NMCC( XI'M #O 9 hhbs!by bbl!bhhsb!!b o ':x/ '2Q'i}i%:,'Mf']Q'LOhy%f,^^'; 'lQ';Oji5f'X,l m o N r-y

s. vmxwv - v x wymv-v s vmmy - s E

NX / \\- MA93MNX / \\ X4XO3MNX / \\ X4X'XYWM / u w ymqxy9w wwxxqw wwxygw z M#x%/xD<NM#x%/xl<NM42W/xMM NMYC<l>CN'$$3L%YCC'\\'lx%Y)<l>IM 'O ~ 9 N 3! N ^ ' '50I'll'$NYO?/l'ZT'5NC',^^'s"'}Q'$N%C',^^'/,' o ~ a d g,v . vunwm v s vmwu-v s vmx s wr-w // \\ MX9:Tn\\'4\\ / \\ MMTWsX / \\ MAOWsX / w >m&xx9w >m&xx9m >mqxxvw a M f/ 4/xy MM&/x2<NMM&/xYM ~ S MY9<l>INM)CNM69(S>CM/ E - bhhsb!!Nhhhb!!bhhkbs!b g x by(sgjj g ;xj9 ; g g j'xjNy;s$3jjyy,'x,f o j e - v x wv,c<wr-y s vexm - v s vmwu - s l u m ixw>xnw / \\ wmrux / \\ wXuxw / w wwxxgw >m&xxwxxx >m9txx+w >lN%##A/xys M

1. xWxrRNNN#AWXMMO

\\MY)<$>(NMY)CNNY)<$>IM k h h k h N 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS RUN 900 Figure 2A.20 Caparison of response spectra for foundation signals obtained from the inpulse filter (solgd line) and the zero-phase shift filter (dashed line). 'Ihe 180 cmponent of the October 16, 1979 accele-quau was filtered with the radial Auxiliary Building /accelerograph pad transfer function for 'IVst 4, shot 4.

D D \\ \\ O N s o-b b .: ic\\ x'; ' :w,i,'x /N,;'a' 'D7, vx '. '/. ;N: W,c,'x / o ^ A i s v uu - Ain s suum / ^ is u, /u / ^ a //Y \\ Y/VA XYtV / Y \\ Y/VX XWY / Y N YNX XVtY /3 /\\ I X/Nx XWhX / \\ X/NX YWhX / \\ X/NXXWhX / W >P,%fXX%'X >/MXXMNNY >WW XhKNNX X\\%#/yM/,X{ hM4//M/X )Chh%#/,%/X)%\\'\\\\ fo NM3NCNE/000sWE000xW o e. d%b!!b hbjMMbsb o m o , m ,'\\ x'c c;xinf,'XA^x x cxw//x Nx'a'x;nvx- / ^, ^ \\ suuw - A \\ www Ain \\vww ^ //V X VNXXW/ / Y \\ YNX XW / / Y \\ YN2 4YW / 3 E N / \\ X44/X XnsX / \\ X4N XX%sX / \\ X/NX XAbX / o w \\,;<fy;,yypw ymxxpw yyysyxhex xx%///PX/XXA%#A%/X 'F%%#4X/X X M% z NMkNxM3)OONY MNCNM O ~ E EMnbaR$$nCiM$tsCr / / / (V /1 \\ \\ VJ'49 I h/\\ / \\ /'. \\ \\ 3(J XV / / /\\ /. s. A NN:' <// I /\\/ /'e s II 's.r\\ _\\ N: <// Th\\/ T.'N \\X LX// /\\/ N \\ fi \\ \\P/jL'XM I'/ /\\ / C / A/i A I\\ \\NTMYN1 t /\\ ^ ^ \\ \\/VXXYx/ - ^ A i\\ \\N)' XVNL/ / ^ / //Y \\ YW)'Y%Y R/"( \\ VNXXVV / Y \\ V/VXXW/ / 3 d W /, \\ X/h'YX1AhX \\ / A . X/A'XXWNX / \\ X/W"XX1ANX / T vx/ W/Mxw&W < DWMm>W >EvXxh>w 4\\%,/M%/X'FM%/MXXNh%/4%/X[)C )(k% s o S A M,3000R M3000xW3000xM 1 )! k h u-A o x A X'c N ': ;nf 'X /N X'\\ \\ex;nt/x / \\ x 'c a% /x, / - A A i \\ swou ^. A i \\ ss/xu / ^ e \\ suuw / ^ / / Y N Y/,/X XW N/ / Y \\ Y N X)Ysy / Y'^_ YNAXWV / 3 sv / \\ I X///XXM4 X / \\ X/N XXW / \\ X/NXXWsX / l W W/X/XXFMW W/WXXhK\\\\X W/MXX%\\X l >h%#//X/Xi>% % V4%/X X'b)D s %# M/XXh4 O sM/000x's2Msti(x300063000xM l kMVbhhMC!N'hsC#I 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS BUN 837 RUXILIRRY BUILDING Camparison of Auxiliary Building envelope response spectra of signals Figure 2A.21 obtained for the 50th percentile spectral rnodulus ratio with the 2npulse filter (solid line) and the zero-phase shift filter (dashed line).

e w w w .f N 5 g. k k 8 ', 4,'i'2K'$iiE9,'\\'/ ', ' ^s'Sd$ ','X "'J'\\$ dsE!X/, / y x vemwv-v s vexwe-y x vwmv-W /\\ MA'X T 6W / \\ MWXTANX / \\ Mh0DKhhhX / ~ w xw<xx9W Xuqxx+w wx9xxN><xy A%#X%/XX&%d%/XM% %#XM/XM%% y NM69OONM@69<l>CNMY9<l>CNM O bhhhb!f hkb!!Nhsb!I go ., x,x]Qxg g7'xj~,x'~;'jgy,,'x,s j Q ~ g g ', y,' x,, en o N c - v x vy> owe-v x ve m - y x wymv-s E W / \\ MMXYnNX / \\ MX'XT6%X / \\ MAO fnW / u w y,fyymxw yxqxx9m yxqxxqw z %%#AK/h%A%A%/XXA4#X%/XXA% NM6901M%6COCNM6COCNM 'O ~ "+ $o b$9mb;!ES$nb;M inC;E 'xAX'c "=,AX A'M %'x ^x n'M 'x - o , un swaw .a x waw ^ ^ s waw - ^ ( - v \\ VA/>DW/ f \\N. VNXXW/ / V x VA(>SW/- d M r/ \\ MXODC6%X / \\\\N MMDCANX / \\ MMODCENX /

y w

xez9xx9m w:x9m XM9xx9m o AwawxXA%#x wxXA%A%/xXA% S NM690CNM$6COCNM8690CNM /

• E~ bhbhb?bhhkb!Mhhsb3!M

' */I'25'&N$ '?l'll'S5id'?i lN'?555*,' v x vwmv-v x vemv v x swxmv-, W / \\ X/h'X)CAM / \\ XdA93CANX / \\ 1MW'XX'AW / w X'A9xx9m X'Axx9m wx9xx9m A%#/F)</XXA%#/JN/XM%NNNMWXMA% O NM6CO(NME00CNM600CNM "+ ! M nCvR$1sC:MMinCv2 0.01 0.1

1. 0

. 10.0 PERIOD IN SECONDS RUN 838 DIESEL GENERATOR BLDG I Figure 2A.22 Cmparison of Diesel Generator Building envelope response spectra of signals obtained for the 50th percentile spectral nodulus ratio with the inpulse filter (solid line) and the zero-phase shift filter (dashed line).

7 e s 5 s o. k Mk k k ) $ NA ','X I',M' $ 5' 'X ', @ 'r,Xf'2D$$$'!X/VD / s macwv-v xvuxxwy - - y s veme-y s v M / \\ M7/XX'6%X / \\ X/h930&W / \\lM7#XYAbX./ w WA9xxs?w WA9xxs?w WA9xx4xx h %///& /xX;h%#7x/xXRQ%N %%///%/xX s NM69(S>{\\M69<l>CNMY9(S>IM O bhhM( )!bthI))!bhhhC)!b o e y i 'JDM$,'XfNJ}$ip'h,'Xg v) o /X "' X'cj}$ 'X N r-v x veme-v x veme-v x wnowe-, 1 NX / \\ MX9306%X / \\ X/hS306NX / \\ M7/>D0AsX / o w WA9xxs?w VA9xxt>w wxOxxmNxA% h'4&%'/X X A%//A%/VD<Eh'4///%/XXs ' s z NM68(S>OM8Y9(l>CNM69(S>CM 'O ~ d! d! U o ' ^/ X^Js'sM- ', ^ ^ ' ', ^C SF4 ',^,' ', ^CBSR"4' ^/ ~ J r ur x vmwy - y s vmwy v x wnawe-3 d sX // 'A MXTTANX / \\ MXO3OnsX / \\ MXOSCANX / T wx WAvxXs>w WA9)xs>w WA9xxs?w XN%//M/XXsh%//M/xX{h% %///%/xX s e S i Y)CMMY9<l>CM86)<2)( M / "+ E~ NhhMb!!MS$nC)!M8tsC2!M ,:' $fQgjg','XjNJcgfgg,'Xj ,'xjN,'-{x3g o j veme-y s vurrwv - s r-v x vmwv-v s.MX92CX*X / \\ 'MATXWW / M / \\ Mnf>3OXsX / \\ w WA9xXs?w WA9xWW WA+Yx*NNX A%//#,x/XXsh%//M/XXsh%//M/x3Gh% o NM39fM869(S>INM69<}CM Nhhkb!!M8 hnC):R$ hkb!!M E 0.01 0.1

1. 0 10.O PERIOD IN SECONDS RUN 839 SERVICE HRTER Conparison of Service Water Punphouse envelope response spectra of Figure 2A.23 signals obtained for the 50th percentile spectral nodulus ratio with the inpulse filter (solid line) and the zero-phase shift filter (dashed line).

. Question 3. Justify the statemsnt on page 28 of Appendix B that "lognormal statistics are appropriate for use here"; document the actual distribu tion of ratios about the mean. Explain why spectral ratios are averaged geometrically in Appendix B and arithme-tically in Appendix A. Justify or modify the estimated reduction factors in light of your answer. The physical processes which affect wave amplitudes in foundations or on ~ instrument pads as compared to those in the free field on the saprolite surf ace are expected to be multiplicative in nature, rather than additive. As a result, we expect that the distribution of the ratio of foundation motion or pad motion to free-field motion will be lognormally distributed (or, stated i another way, that the logariche of the ratio will be normally distributed). If the coefficient of variation of the ratio is small, either a lognormal or normal distribution will fit the data adequately; it is only when the coeffi-cient of variation is large that a difference in distribution fit will be d obviour.. In the case of the soil pad interaction study (Appendix A), the coefficient of variation of the pad / free-field spectral ratio is small at each.f requency. Thus the average reduction f actors can be computed either arithmetically or geometrically, and similar results will be obtained. This is illustrated in Figure 3.1, which shows average spectral ratios computed by each method. For explosion test data (Appendix B),the justification of a lognormal distri-fution for the foundation / free-field motion can be examined from the experi-mental spectral modulus ratios obtained. Figure 3.2a shows observed values of the logarithm of Auxiliary Building / free-field spectral modulus at 25 Hz for Test 3, plotted on normal probability scale. There are 40 data available for this test series (5 shots, 4 free-field sites, 2 horizontal components). Also shown is a line indicating the fitted distribution. The linearity of the data, and the good fit to the straight line, demonstrate that the lognormal assumption is appropriate. For comparison, Figure 3.2b shows the same data 3.1

but plotted as the arithmetic ratio (not the logarithm) on normal probability scale. The non-linearity of the data on this plot show that a normal distri-bution is inappropriate. Figures 3.3 and 3.4 show equivalent plots of data-for Test 3 at 14.8 Hz and 30 Hz, respe ctively. While the data are not as well behaved as at 25 Hz, the lognormal assumption is still the appropriate distribution choice. For the Test 4 series, there are 16 ratios of Auxiliary Building / free-field available at each frequency. Plots of these data in logarithmic form on a normal probability scale are shown in F13ures 3.5, 3.6, and 3.7 for 25,14.8 and 30 Hz, respectively. The data do not fit a lognormal distribution exactly, in particular because of the fewer number of points available, but the assump-tion of a lognormal distribution is justified, since there are no systematic departures from the fitted distribution over all data sets. l l t { [ 3.2

AVERAGED SPECTRAL RATIO FOR HOMZONTAL COMPONENTE 8 E' \\ ~ Ei g. ~3 3i h-2< "d" ~ s 0 0.O 10'.0 2d.0 3d.0 ed.0 50'.0 6d.0 70.0 Freeuoney. hr cEourrme AMITHMETIC Figure 3.1 ' FREE FIELD / PAD SPECTRAL R ATIO 4 COMPUTED GEOMETRICALLY AND ARITHMETIC ALLY w -w-- a-- e-- --r-w- -22 ,,_,, - -yr ' - =, ,w-w-p *,--' -- - -p w y-9 9 --FT-y p" --e'

.I I Je e .98 - e .4 - e e .7 - 5 e 1, .s - e e ? e .s - = e j e g a-e e .2 - e .3 = o e .0 $ - e .02 ( - 8.0 .S .S .4 s2 0 .2 4 LCg(AB/FF) at25hz l i { Figure 3.2a LOG OF AUXILIARY SUILDINGA'REE-FIELD SPECTRAL MODULUS (25 HZ), TEST 3, PLOTTED ON NORMAL PROBABILITY SCALE l m

'q o s t Jg .99 = e e .9

• e e

e .8 = e e e 7* E e ^ e .E .S = e c 2- .t 5 ,e

• 3 A=

t a-o l e' t I e e J* e e .05 - e e

• M s

u a O .2 4 A J l.O S.2 g.4 AS/FF at 25hz Figure 3.2b l AUXILIARY BUILDING / FREE-FIELD SPECTRAL MODULUS (25 HZ), 'i TEST 3, PLOTTED ON NORMAL ~ PROBABILITY SCALE

e 7 I E f 9 e .S-e e .8 - e p .7 - = t .s - e e* a-e .I - e e e .t - o .0 5 - o e .02 -LC .0 -A -4 .2 0 A .5 Log (AB/FF) At 14.8 hz Figure 3.3 LOG OF AUXILIARY BUILDING / FREE-FIELD SPECTRAL MODULUS (14.8 hz), TEST 3, PLOTTED ON NORMAL PROBABILITY PAPER

e .9 5 = e .9 - e e .8 * = e e E o' O 2 .s - E-i ] a e 2-o e .2 - e' e e .I - e .0 5 = e W> .02 =9 =A =& =A =.2 0 .2 .4 Lag M) At 30 hr Figure 3.4 LOG OF AUXILIARY BUILDING / FREE FIELD SPECTRAL MODULUS (30 hz), TEST 3, PLOTTED ON NORMAL PROBABILITY Sd LE

i as - t .e - m.r - = 2 j.s - t s-I i

1. ~

t-o a-s/ / l s. .os - .0E ..e 4 o e Log (AS/FF) ot25 hz 1 l Figure 3.5 l LOG OF AUXILIARY BUILDING / FREE-FIELD SPECTRAL MODULUS (25 HZ), TEST 4, PLOTTED ON NORMAL PROBABILITY SCALE 1 l l

.e s -f .s s - j .9 - e g3-e z .l .s - E .s - fe s-e e 3 2-e .3 - .e. [ .o s - I .o b -L2 .t. 3 .g g Log (A8/FT) At 14.8 nr 1 Figure 3.6 LOG OF AUXILIARY BUILDING / FREE-FIELD SPECTRAL MODULUS (14.8 hz), TEST 4, PLOTTED ON NORMAL PROBABILITY SCALE i [-

l i i ,, e ? .9 5 - .s - e .9* b .7 = = 5 n .s - =j,- a e .2 - o .I

• e

.0 5 -

• M i

-L2 -LO -A .s -4 2 0 2 4 Log (AB/FF) At 30 hz Figure 3.7 LOG OF AUXILIARY BUILDING / FREE-FIELD SPECTRAL MODULUS (30 hz), TEST 4, PLOTTED ON ' NORM AL PROBABILITY SCALE

Qu2stion 3A. Th2 Licanese should documtnt the actual distribution of sp2ctral ("# O' 1) ratios and justify empirically the use of legnormal statistics. Based on the actual distribution, the Licensee should calculate foundation / free-field spectral ratios for percent 11es ranging from, at least, the 50th through the 90th, in 10-percentile steps, for the total site-to-site and shot-to-shot variability. Reduced envelope spectra corresponding to these percentiles should be calculated, using the results from the pad-shaking teste. Distribution of spectral modulus ratios The distribution of spectral modulus ratios is documented in the response to Question 3, and it is shown that the lognormal distribution adequately repre-sents the data. Discussion of the response spectrum criterion The calculation and proper interpretation of reduced (foundation) envelope i spectra corresponding to various percentiles requires consideration of what the original envelope spectrum represents. In particular, if the envelope spectrum represents ground motion at a specific site, percentiles of founda-tion response should not be developed as though the envelope spectrum were applicable to an unknown, randomly-chosen site. Hence a discussion of the possible interpretations of the envelope criterion is appropriate. The criterion specified in the Safety Evaluation Report, SER, (NUREG-0717, Supplement No. 4) is the envelope of response spectra of accelerograms re-corded at the USGS accelerograph site on the dam abutment. The criterion can be interpreted in different ways that may or may not be mutually exclusive. The most direct interpretation is that (1) the criterion is specific to the USGS accelerograph site, where the strong motion data were recorded. Another interpretation that is consistent with the language cf the SER is that (2) the criterion is referenced to representative site conditions, with the implication that i 3A.1 -e-- .m., p.- ,.-.,,--m - - ~ - - y

(a) the dam abutment site is representative of an average free-field site, or further that (b) free-field sites are indistinguishable in terms of seismic response. \\ These interpretations of the criterion were taken into account in the design of the 1982 explosion tests. The outcome is that, while there is substantial variability in site response, the dam abutment accelerograph site is in fact representative of the instrumented free-field sites in that it exhibits average site response. This is docum:sted in Appendix B, and is discussed at length below. Thus interpretations /* and (2a) are equivalent for practical purposes, while interpretation (2b) is not supported by the data. The question being address,ed implies yet a third interpretation of the criter-ion, namely that i (3) the criterion is referenced to an unspecified site chosen at random. This interpretation is neither stated nor implied in the SER, and was not considered in the design of the explosion tests. Becacse both the data forming the criterion and the explosion test data are keyed to a common reference site, variability of site response at different locations should not contribute to the experimental uncertainty in estimating foundation response corresponding to the criterion. l Results for individual recording sites l To illustrate the importance of the site effect, and therefore the need to apply the criterion in a site-specific manner, results obtained by applying the envelope response spectrum criterion to individual recording sites are given below. These calculations demonstrate that the criterion from the dam abutment accelerograph data cannot be applied indiscriminately to other sites l without regard to differences in site response. Application of the criterion in this manner introduces extraneous dispersion in the calculated foundation spectra that would not appear if site differences were taken into account. l 3A.2 l l

For each site, a zero phase shif t transfer function was obtained from the 50th percentile foundation / site spectral modulus ratio for the shot sequence. For both the-Auxiliary and Diesel Generator Buildings, Tests 3 and 4 provide a total of seven transfer functions. For the six free-field sites, foundation accelerograms were calculated by filtering the ERIEC free-field accelerograms computed for the October 16, 1979 RlS event, while for the accelerograph site (P2), the USGS accelerograms were filtered to produce foundation accelerograms directly. In the case of the Service Water Pumphouse, the ERIEC free-field accelerograms were filtered with the four transfer functions from Test 3. 1 Results for the Auxiliary Building, Diesel Generator Building, and Service Water Pumphouse are shown in Figures 3A.1, 3A.2 and 3A.3. The calculated foundation response spectra differ significantly, depending on which site is chosen as that to which the envelope response spectrum criterion is applied. The actual foundation response measured in the explosion tests is of course not a function of free-field site location. The scatter in the computed i l foundation response spectra illustrates the extraneous dispersion that is introduced when the criterion is applied without regard to differences in site response. Note that the amplitude of the calculated foundation response is inversely j related to the spectral amplitude of the signal recorded in the field. Thus the site with the lowest signal emplitude, F1, has the highest calculated foundation response. But if the accelerograph pad had been located at site F1, the envelope response spectrum criterion would have been much lower. To examine site differences in further detail, shot sequence statistics were calculated for Auxiliary Building / site spectral modulus ratics for sites F1 ' and P2 (the USGS secelerograph site). Figure 3A.4 compares Auxiliary Building foundation response spectra calculated for the 16th and 84th per-centiles of the spectral modulus ratios for F1 and P2. The separation of the results for the two sites demonstrates the significance of the site effect. Although there may be a contribution due to path differences, this is a relatively small ef f ect as can be seen by comparing Figures VI.C.47 and V1.C.48 of Appendix B, which show very similar Auxiliary Building / dam abutment spectral modulus ratios for Tests 4 and 1. 3A.3

Because saprolite propagation effects are site-dependent, it is insppropriate to apply the criterion from the dam abutment data to other sites without regard to site dif ferences. However, comparison of the results for different sites in Figures 3A.1 and 3A.2 shows that the USGS accelerograph site is representative of an average of free-field sites in terms of site response. This can also be seen by comparing the results for the accelerograph site (P2) in Figures 3A.1 and 3A.2 with the 50th percentile results in Figures 3A.5 and 3A.6, respectively. In the latter two figures, the criterion is referenced to the geometric average of horizontal component spectral amplitudes at equidistant free-field sites. This comparison shows that results obtained by referencing the criterion specifically to the accelerograph site (interpreta-tion (1)) are nearly the same as when the criterion is referenced to average free-field site conditions (interpretation (2a)). Interpretation (1) yields calculated foundation response spectra that are slightly lower than for interpretation (2a) for frequencies up to 30 Hz, and slightly higher at higher frequencies. To a good approximation, the accelerograph site can be taken as representative of an average free-field site, thus validating interpretation (2a). This finding is consistent with the observation that the saprolite thickness of 56 f t at the accelerograph site is representative of the average i saprolite thickness in the Monticello Reservoir region. Appropriate percentiles for reduced envelope spectra Consistent with the above observations, it is appropriate to calculate per-centiles of foundation motion incorporating only source and path variability. This corresponds to interpretation (2a) above, which is equivalent to inter-pretation (1). Figures 3A.5, 3A.6 and 3A.7 show percentiles for reduced envelope spectra in the Auxiliary Building, Diesel Generator Building, and Service Water Pumphouse, respectively. These response spectra were calculated from accelerograms obtained by filtering the free-field accelerograms derived for the October 16, 1979 RIS event from the pad-shaking tests with zero phase transfer functions of percentiles 50, 60, 70, 80, and 90. Data for Tests 3 and 4 were given equal weight in computing the spectral modulus ratio statis-i tics for the Auxiliary and Diesel Generator Buildings. It is the Licensee's 3A.4

position that Figures 3A.5, 3A.6 and 3A.7 are the appropriate refresentation of dispersion for reduced envelope spectra. Percentiles for reduced envelope spoetra according to interpretation (3) The overall dispersion of the foundation response spectra shown in Figures 3A.8, 3A.9 and 3A.10 expresses the uncertainty that would apply if the loca-tion of the USGS accelerograph site were unknown, corresponding to interpre-tation (3) which is inappropriate. The results in Figures 3A.8, 3A.9 and 3A.10 do not apply to the criterion and are included only to complete the response to Question 3A as stated. l l t t i s 3A.5 t.

s@ o N N s o-h's I' [ x g . n :-; ~ :L.:. , r: :,,,,-: n., s;'a.. :.L ', : a - .n,, o , e, /, i s .w., e.. 4 4 ;eis rs, u m 4 ,. i A, s u, o.- , o o e f y y yexvwy f v 3 v...xyvcy, x y.yy 3.,., y f-3 'A'f..S.I'fi f *. M s\\' / 's l M4.* M ',t'b % X /' W /4\\ { >' /) '.4 'i1.MQ.' / / f xxx' i x' xemxcex x',t. <.exsw y,w>rxNesx' X\\MMf'Id/.XI l'.TNO\\b / M d 4 / h I M 'f.k N b /,I'M/N ) il k %s% ./ y \\MSkd'/'Il'k[l'N$$d'D'd 3'I!IN Y N Id'E'/ k NiiN Y d O S N -ty,, n :. c F1 gs...u.1_ : -w: w,L '>',,. :.a :v,4 :,;, x, w a i vs ~ ,..,s , s s,mu,, ^,s uwu-A \\ /,y N l sy\\ y'Q'f'j y f/ N./pN'ay%%/ / %? \\ y yNggyy / \\ r y ,xw ex / x xymwx / s wm,ux, U NNY )f.'P3 WNNN( X/ X/>('KshN'X' X/,'X,iM XNANX' / X\\\\'\\l#$'.YNg DMN//4X/X MR'%#//M/M M 'f>'N' 4N \\ z s s'N'MA//Mh\\M6CO'$ pdm @Z:ONx'M 'O ~ /::

x bMm"A Ac ma,'x, a >, : ' :4.

'X, o // Nr U 'A N A/tOA ' 1% AL A\\i \\ 4 FR ' ' ^b'V"##M ' ^ f// //// ft N /vYX'P.'<>1Y %\\ NWV '/ N' \\ 'V,'iX VNN' r' / ) & Vd/ IX/ADOt'5AM \\/f%W7M' F1 / N g ywxygsy / W T W//5/ X@X'xNA%1 \\MW/ $ pg y NSX W/MX *NnNX o M if/ M A WX K'Fl e h W @ Fs,P2 g g'g 77ppXj \\\\\\ S y 'df//@t%YCOU'NM69O(N'M'MO@M / E 'hh b b) Ihhbb(2k o /,s,n ' :. sRm'; ca,N x '..s >,c c .(, s x.n/.m, :/,s' , ^i^. x x.v. xw., #. A n nwav. 4 - s

1. s. r m.n.,

- A ( / 'Y N 4 //O4'W / N' \\ 4 fxji^tY / M NP A ' W v.'s vr / ) NX /l\\ l X-f.N'Y'dAhX /.\\ { X4Afi'MhW /\\ l 'X'#?s'4 'XNANs' / s >',' ',l/M 'Xh{r'OD'( M///vfd'id'fINhX' M'/X4'>4 'DXN\\'< NNX' I / l ANW43WMMiMW##,</b1'8N4MNi#MM/M MtRx'% o NM$k'{',5)( XN M M/C(XD'xN$d/^k MM ' / 'hh['bsb!! I'b?[ 0.01 0.1

1. 0 10.0 l

PERIOD IN SECONDS RUN 823 -ENVELOPE 728+729-AB i Figure 3A.1 Auxiliary Building envelope response spectra obtained by filtering free-field ac lerograms frcm the pad-shaking tests with zero phase L shift transfer functions for free-field sites F1, FR, F3, F5, T6, l and P3. For the accelerograph site (P2, heavier line), the USGS accelerograms were filtered. l

o o c o s s. o s o-s b 9 /x' N /. .+ a, a ex'aca.u x a',c:: /,' x & x'c x ;:',:- o

v. we - s e ixn. w m -

.s n.w n. - - s N V/vN / -l ' x y >S X X %', y /h O / x winm'4NN.'v in wimvw / / y' N 4',y'x xv4y' /N ~ ,X /> x rwm,.a / w i w/x4 m xxxx w.m<mxsw i mux xdxw< XNNNN//AM/x!NtXM/ZAN/MXMhE4W4x%\\\\\\ f NM31$'/9<IX'NM6I'NXPs!NMMid/7NN's#A O b$$,$* Y.h0x R o ,. z a, A, 3u':P " ~ =++.':n,w=>w,,'x,. x'x.caz,yv,t m o ^ ,,3 3,s aqq, f y; 3 i s s. ,y y N r/ y x/t - s </ y x NwxYwv / v x verewy / h WJA/_ Ti ' / \\ X4N f DAbX / \\ !XJ>fX fMshX / ynggf4m'K E NX /\\ v >wx4xwxw y,xux>xssx o wMNNWW/N ih%&Wx'MVXxN'N#4X/W X'iM'& z h's)N1M'/X(NM O NYY8% NMh ~ $9 NNN MO Nb[ X-x X V 'D; N',f'X - x /r Kw',v;'mtref=%'D ra o A ^ i s N A 'w w 14 - A /s ters a N rxu AAiAm wp1 gr/ ^ v,5YxNw w n

• W F6 / Y N Y'YN VY'T / 3 a

F1 V W sX#N/W X'XG >NW \\Y IT h 6'r', F3

1. \\ *###

/ P3 T p3 F6 M /fl yeyggMWf)P2 F5 'X Y'E4DNY @[NhM N TN M b W MX NNXbbbb/8M/N NisDN b FR a P2 /j' 's\\ /7XK's'916!$f2Xl>('ixW / F3 s d, hbA hk(F d E-1 'x n A ' n ; w'x, x s:.. W ' >: x :'cn.u: t \\ o -/ A > A 4 \\ h A/* siVMA /^ A \\

h. Af t TVu ' / A A8\\ IN A/4 M VV.i /^

N N 'ON MNN '*" /Y X N'<'r'X M NN 4' / h ( / V N T',N T d N*t 4 /' i sX /f\\ l XJ/,N XNW / \\ XAM.UYhX /lN ! XAAf M X'Md'X / Wl w/X4xKM>CNNX 'h'/XMXhKW i w/MX XbKsNX MNN4SM/}X MNMN //MM/M M'Is%sMrM'lOM\\\\\\ O \\ i \\ N M N M $ $ M ' D 's{ N 'N \\ M ' / f N M N '/ N )>( 7 m / 3 w . ;Y p;- o, \\ /\\ /y y /8 Y $ '/ W R Y N / ss x. l 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS RUN 824

-ENVELOPE 751+752-GS Diesel Generator Biilr % g envelope response spectra obtained by fil-Figure 3A.2 tering free-field accelerograms fran the pad-shahng tests with zero phase shift transfer functions for free-field sites F1, FR, F3, FS, F6, and P3. For the accelerograph site (P2, heavier line), the USGS accelerograms were filtered. i

gr~ s 9/ e e so s s s e A N \\ ) . I! s 'sl sY 4. J ' s , ?O sl ? R4 $.i ii 'lx., ?' '.st 2 M 2..., [<n' O 4 ets ,i f A. \\, _ v5 g v M / ,s, A oA, 3 ( u.: / A Q / A 'A Nex ev r < + 1A / Ae r / ,Y N Y d' N Y Yn' / 'y~ \\ Y,YYXV,y / V N V A ' 4 <y6 - / ) NY / \\ ! */A ^ [ 'A_'t 1<* M / \\ M.X[X l'$MX / I \\ f )r'AMf 4 )L'WhX / ~ NT W/h<f.D'hMNM 'X/4'/X'tM(NN)( l X//M'>f XHTNNX >OIWNk V/iW/'. W % b k Q %# 4 M /X'%%'NN y >N%O/VA/ N NMEd')C<l4'I\\MlMM >tNMh%y)('X.'M O 3dM $$ Mdh(hh 9 u ' L' &, ; ', ,s N ',' n;Q..;;,l', u> O 'e' X ! s -], i F6 ,'J_'._f, A ' m / AI A i ' ' Wi1 ir, / A A s \\ A< CO.1,1 / A A IN N A/YIe / A ( / Y ('Al eM 4 ' / f X.,.YN NYtV / V '4 V,*V's NYtY f 3 W / \\ ['dM/hX / \\ ' X/NXXMNX / \\ IX4NYXtAM / U SW M6Mf4 YAMNV aX'/KWXMNX W/hXXhMNNX >M\\%V/d>444fdRSS'M'dX/xhN>S %4 9:/X'XS(4% z NW4'd8QMYb '" WM4'))<XM 'O ~ / /1 [IsI N \\_N)-N,$ $ ISY,' x)N s' b v'), Ie x )[xYNWI / ex' / Q M / / / N2 /WA \\ \\AS. MtA / \\/N A \\ ' vVx W - 1 / A A iX N A>i W P. / A ( / f/V //k Y/VXYYh' / V V V/FXXi4Y / V 't \\'.VX XV{Y / 3 W W47 ! 'l/ 'XANYY M K X' / x 'X'WXYWX / \\ i X/N'X)'uhX / T\\ X W/X/X XhM\\W M/_U/W/X4X W\\W W/MX Xh N r1 w ex/x?c>R%ae/x kN>mvsx/x x

• ks O

(T) ( ) \\W F6 / E -$NN)![ 'b5Mbsk O j,fixs s' vNg, e'f(,k N xi C h 'g f ) [e ' ,',s' f ) x '<,' [x'hke<)I,'N / ^' y 4'e N' v -.d NV<W / A A I / ? N / A+A si \\ N 4 W _. / A A !\\ s A 1 (V.M V / h 'N Vf, 'y (viv / 9 N l ( / y X V N TA' h n' / N e N /4\\ , 'X.o A) '.( N:'4AbX /\\ l Y/ AfX ',febX /l\\ !X4AfX",tlAhX / ~ xxx w/wxwww W/wxx>x\\sx i w/uxec\\w l >(k%##X/N %K%N67M'AxVx^'\\'N ENA!Xl%%8% O N NMY9< 'X N N N[/CdNM N D)O'Nd '/ hh N)Mhhhd NM E O.01 0.1

1. 0 10.O PERIDO IN SECONOS RUN 822

-ENVELOPE 705+706-HP Figure 3A.3 Service Water Punphouse envelope response aw.i.sa obtained by filtering free-field armleiwi== frun the pad-shaking tests with zero phase shift transfer functions for free-field sites F1, FR, F3, and F6.

h O O N N g g N g 7 y 73 7 yq . h \\ ,/ ./ I .\\ \\ .N /.y N xNh5 Y, ~ e' e) x )<bx)(h,'N Ie'N, [ w s N b ) 'e o s ,w / ,s i Ni(VJ' / ^l AtN h N) TV / / AI A \\ NN)r (V J 4 /A e A A iN N Q ( ./ 5 % Y.N'M X Wi/ / Y Y N% 'X'XNNN ' / Y N VA"N1%Ny /) \\Y /.\\ l \\ *,V Y h'\\'Qs\\( / \\ Xd N 'X'h' 9 Qs%( / \\ X47[X XbDX / ' : :, :vxn av xzuxwwx x//Wxwmv; Nw R@k.,W;</XF}$NbRMk +/9X /XiXMN%/,s7)</>4 )<NKNNNN j# )d\\ MEM@OCNMW/COIM O 'N' s'$MMI'IX h'MONMbhM hbI 9 u . n c, 2;a:y ',:,'x /. s xs'ch,'x Axx)cn,4,x, n o x.n g w - ^, a x n,mu / Ai A i x x4w < u, /s ,s - 3 / / N/ Y.Y X Y NN N' / V N Y /vN X nN' / Y \\ VN)7 %y / ) \\ ~ NM / '\\ /s W[XbAhX /l\\ l XdMYXWsV /\\ XIMDWhX / Eu xw i ymgwww i waa>osgsy w,wxwxsw >Ds\\s\\\\VhlfMD4M\\K^s%Nj#,d%/}XMNSNN////)</X )(its^s\\\\\\ z N2Y7/4CMMMW/90E'N$$$W/90ds'$$@ 'O 'N ~ d S u i ;', ::;'c:,',M % x'ss'My,,'x L x'd,'ana'u - o // I/ A e\\ N NYV M \\4 ^l A iN \\N'Q M / ^6 ^ s NA'Y<M / A e // /V N N'A'M XNNN T Y h % 4W'XXYiN' / Y \\ V //X XYsY / 3 N / /1\\ lX///X XWAM \\/. \\\\ l >XNXX1.NNX /\\ X/,N'X X3AhX' / W Y x c</ i W YvXXhhv\\. Why/xmhNX WX9xXhhxX MNW/PA/XYMW/X4MfKX sXNSNN///P)4/XX'RN% N o s73M's46)fC)('N$P8 MX)Cx*$3[8W/%>CN3$j@ / 8 c,, y , \\y w >' / d ) hb M). IsOCN . AN I W ~ i , x A ;N 0X17J 'x - s x'n'M5' ' f/x / s X'0 s'0 'M', f/X - ,' ',s o ,A,s w w u - A ^ x wY<w / ^ A i x sav<w / ^ // V N YhNX%5// Y \\ VA)3NNY / V \\ Yh'N XYW / 3 sN' /?\\ I X/,NV)f(9Q X / l\\ ' X-NYX9sbX / \\ X4MX'*'1AhX / NNX i W/WXXhKNX WZ@MM\\X W/hY Xhh\\X \\ RsN'N##/'/>4 ><N>6\\\\\\//Z7)</)< )<%%//AM/X )CM \\\\\\ O NM N'2'X1NN N)OUNNN)< NINM '/ hkN ' N Mb v G 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS RUN 806 RUXILIARY BUILDING Figure 3A.4 Cceparison of calculated Auxiliary Builch.ng respcnse spectra obta.ined for the 84th and 16th percentile spectral nodulus ratios for site F1 (upper two traces) and site P2 (lower two traces).

8 s' o-9 sh h k , [s .. '. : n. w,c . x y:v:r ;p :: L.< u :2, ; : a .s ..s.

g., <.,. m,

x .z, N - / v s v w.a.y -,, a n.,u.w w o j.~ x - r +w, s s ,v s s /%s s,.- - s.. /tN l ',c,5N N( 0.4y / s X.irg y nKKv_ f i N Fg<^r.i n,q.x / Nsx i wyxco N(+x v ts m N<+x i X s p a n w >e hN'W 4 M ^ M O ftb 8 h / M N e>Dtp;S%!4MM/?Of%NNN j MN%}'7X >iS.NNMX,IX'i's's9$bE8}'I/ <l'$iM O ^ rg.wXX O, $Nhk' $.[ , x pp - Ws$N4 .'N y gNg / N y Ykb. V Y hY $N '$h y w a , a a '. : 'ww\\. .1.k.'-< ' '-m,m. 2,,. :'.m ;m,s x ; Ai ^ v m asae,Ai + i x u.ww / ^ ^ i x n.v xvm - A r / Y \\ / y/vM t'3p s) #/ Ne N V,tVTyet.' / s :f X N,4 N ty,y' /3 /44N' N YhhY / '\\ h4.N 'XTLJ,) M / \\ N6VM 'C.'t,$>( / 1 M /i \\ i U NNX M'.th4%#/%MNN' X//XAX 'KN4\\\\X >'/X 1YXM{\\\\X >$N%A M'$ M4M#/M</X )N'SN'ss/#PA/t< MM N z NWdC'MM$TO)(>Is'$$%8!/,fOCx\\'$j$ 'O ~ R N x A ; c 4 0 s' u n n. 'x Nh: :. A. ' x ,xc ' O.% ; t ~x / o uss a n w.cw s w us.x s ww - .s .s es s.v w w / ^ f

1. 4# N WVX X'sW K\\VA\\ N Y'VV X4N Y / Y N N. WX V'd t / 3 W

@ /F X l X/NN Y1AbX \\MhW104'X'XWX /' \\ lX4A N'iM,bX / W/8/ X'/CN'kN6NW \\KWE 10,ANNX X//X 1M tN<NNX $ffNN/4PX/X)NM#M/X XN>685\\4SM/XYtM'$b o IN 9j$ '\\' / E [8k W )'C(Mxs,@@)OCD'$5M','d< l> k k [ k o ,, < A. 1, N: c=r i A '<' c s' M ^J ' x,:. ' : a '. m:., ',' x / ,s, s e, mn,. ~,.s,s u w w - .x i , - n4 m ,s 9 / N-x N ' A> 1 MNVV / 't N Y A N t'iti' / 3 r / Y N 1.VX WNN' ! '(4NV DANX /4\\ X I.N'X T'vMX /I\\ iX4 N I D O X / /I\\ / N\\X l M','/XrMXNT'\\NC l X//XM'kNDhN,' l X','/K f.OlMNsX N>$$N d:dX/>4 Mis N5hN/%'Pd/l>d XV>S$N\\ O \\l N >SN\\' NOM'X/X X NMNF/)OCMbCO ('$ M N ! M ) 4 M ' / N(h bl~/x q2 O / N 3( z g y /N N f s 7 .' x k b Yk h h8 d R 0.01 0.1

1. 0 10.0 PERIOD IN SECONDS RUN 792 AUXILIARY BUILDING Ainciliary BdWng envelope response spectra calculated for spectral Figure 3A.5 nedtilus ratios of percentiles 50, 60, 70, 80, and 90: statistics en-ca pass source and path effects. Top trace is the envelope response 4.5 RIS spectra.

spectrun criterian, and dashed lines are the SSE and Pg

e e e f o N s s o. N kd h zx b blk. I k)3 $ o .o s,1, 'n ' M: v, : h '<' w :,::. . x'.;'m :n a.,. / .s a +,a ns43 w ins

w. cw

~ .s nu.wwe- 'Y '6 ) 'A 1 VNN Y / N' N. Y.Nf fiNi' /h / / N* N. Y,N' ( 4 W Y /I\\ l X4N N >f9 M*_ - lNG M X'flXtX -/I \\ X4W'>D4fM AX /' / M '.X /,N Xl'Ns\\\\h' l X'di/X XNMNX NNX X..d.W )'l 4% 4M,'^.')a/x 1s,iiNs**vq 6%4,'4'OX}MMN4MVXQs%% y MMIDOMMMC'hskMMh'.-.Ich'M b S h ) X A 'C A' 3/ M : c : ',.six i_v. w,4A - C. n. J X 4 ' " I 'C,M / 's / u> o .s x m,m i,1 A A.A ve f m ~ ~ f / Y __ T ((/r\\\\g V N \\\\ l.' 'Y Y t/i ?4 N N '4 i / '1 N. S',"i N A'1'NN' / h E /i\\ MONfANW , ;\\ l'M N Xii W NX /\\ u X4N 'r YWW / U NNX >79/@UP\\MX l N'/X !'>' Xh :'nX X / h '4 X K'N M X %' \\ % /4' 8 N / W ) @ h ' M 'M 1' X/N h dN% dM%1Xfh%% ^ z M W 8 /)! M M iMl'Ds'X MR8l4('M 'O ~ $o NNk!)hb M:NMMNMN* ,/' uni.2 't :.:n.;', a % NL t , ~ u v ';'nm. ; ' ; c

1. ///3/A lis N A'i Mr A hdTN-Ns 5 'i m ei / >-

A 'N is 4G W.' i / A F ///,'A' F 'Y /r'?. i 1W / 'N ' WS Nn"2 9.'f % M' / '1 N N 'A') win 1 ' / T W W//N Y X4)#'>' 't\\khX / : WAA"MMMX / \\ XAN X NRhX / y xR,: XMv> xww >mf>r ex xew>::shx s X\\%#O"X/X.'XNXM#/@ /XY 4% $$hW///9WXX s o d'NMhb)M M86'D M/ S MNY)[ "+ CsM#Cs! M K M' E~ ,' x A ^'I's x'i m. 'x. ' c s sm . x a ' 's s'ut,c 'x, o I . ^ . x xarma / ^ ^ix x~w e-3 ^ x 14., w e, x '/N' N VN\\CsN N" N' V YsN 'Y')' SNY 3'?VM (1'4 y / 3 5 a / \\ X'M'). K14X /\\ Xi AO 'L'V M / \\ X4NN ;4').\\hX / h\\X X/hW 'KNtW M'/K/x Kh4N'Rs' 'A '/M>: 'xNtNNX A%,4 W x /X XO6'x'%%4Wx!1P& %4WA/)ONb M o M M.'dlOK'sMfM'/$$M'@'@MON M '/ I'

1. +

$'Nd,l k[' T)>[' 2; 0.01 0.1

1. 0 10.O PERIOD IN SECONDS RUN 793 DIESEL GENERATOR BLDG l

Figure 3A.6 Diesel Generator BM laing envelcpe response spectra calculated for spectral modulus ratios of percentiles 50, 60, 70, 80, and 90: stat-ist .; enocripass source and path effects. Top trace is the envelope 4.5 respc.,nse spectrun criterion; and dashed lines are the SSE and Pg RIS spectra, i-l

O O O O o 0 N N g 3 s o. ) ,)k h' ' ' <~ / is ', ' ',J I ', i s 'lO3}li'^^T'A'ic'3 33's !' 'iti ' '(Och!hd',X,/ , v x

s.,s w,s s.

,- v s v..,ws.,4- ,v 4 v,s w s.s,. /3 N' /l\\ I (4 44 'h' MAhM / !\\ lY/ NYO3Qs\\' /4\\ j X.gy %QN' / h

\\w i w,+s.txxiunx i w/xwxwwx xgyxwN*sw M'hN4I'}'X/hdM M N\\N 4d'ySi/MMkNN'4,4M'X/X3M> INN \\\\ /#

MM!IQd36'SM@id$CdiM@^K hiM O 9 b [d4 1 U ( g e a A x'/s' m 's _L':_.u:A ::. x,1 x -0 ' m.;: x / .s / A i A i s/ fs W (Vt A, / A A i s N Wa X 'M / ^l ^ \\ h_&T W U / A w h N V /VX ('ss V / Y \\ T ^VM X3'N 4# / h ' / N' N / N',-S'A XM h E sy / \\ t h'4NX'JiY/Mf / 'N I XA Ny 'r'tQsv fi\\ X4NM*VQW / u Nw w & fx A s % sx wme<xN<sw v & r>>xN(sw V 'fs % % ' O A\\\\\\MM'M/>fA\\'MN44M/X 'kHXNN'4 4M^A/X z )>('M MMMNsYdhd'/D< >IMNSA ~ $9 N kbl Nb$Dx s,s 4 ::,tc'>t, x x'A' 6 ;:.'n / ,s ,.,,.m-,~, ,-,m a - w w. s s wxw- - 3 x.ssuova ,s annews- ^ (/ /IN /d M V/sX 'alNSY / V V 4'N'fX sNs' / N \\ V/VX XNN'V / N W sX/ '/-91r t XO/Yx'tKhX / N X/ W W N< / N-X 4 N 'xWshX / >: r* w &fn N<xv W4fx XhxNw xe fx1MxNX R 1 NN MW X XAN%#WXN'A % 4AWXX'A% o S lnkW/D0('\\W'/WhDW&'ND(l>D$/ g yy j 3 v g73 ,z 7x y c, /~; Y b ' / x' sb b$ MIN { ~ i x. N% / ( o , A x's'sem,m a >c'0Nm . 'x a 'A =g wx,- - - < ^ n sews. A ^ n ~ ya w - ^ ^ sNews- ^ / V N VA YT4 W' / V '4 Y,$'YX'iN s' / N YNN Y sN 4' / h r sX / I \\ l X4 MN~ONN< /' \\ f X4YYXV K'*' / \\ l X4 NX A NNX / Nsx i wAfx1mv xen%v wmxxNMY AN\\N4&/ /bOs'i&%#/rWXM'ihNNN4WN>1'k'shh% o f M W/DOD M W/D< $ s> ?$ W /D < >d'M ' / x /s y /s n c,, y,3 w /s y r . h hy' Y k 1 b A b

0. 01 0.1

1. 0 10.0 l

PERIOD IN SECONDS RUN 791 SERVICE HATER Figure 3A.7 Service Water Pumphouse envelope response spectra calculated for spectral nodulus ratios of percentiles 50, 60, 70, 80, and 90: statistics enccupass source and path effects. 'Ibp trace is the envelope response spectrun criterion, and dashed lines are the SSE and P 4.5 RIS cpectra. 3

o N N s. / 8 j j ) x N / )/ V ' k',hh \\. N\\ / >' / '. b \\. - /(, . l\\. \\\\ N t. O /',' .- ; Y U,,' ),',' $ ' ' ',',. x '- ' s },,), * ', ', ls.,, d,., f, g /Ae a l'. lx A %.f l. ' ! / i' h,v d s. "- -. /. i k N > g.; 4 f A # Y $4 y/r M VYi y' /T Y N 9 N'4 W "4 N T ~ ./ / V '6 N /r'x (O P M / \\ i h'4 N X '8.l Ah,X / \\ l X4VX 'i'id l'X /:\\ ! X4 NM I' MM / \\NX >V/FM C:l N M' '>P'/Ki'A XISDsW t 4/h4MKfC>M /W/4/ 4 X'EX R % M @ '/ VXYi? M M M E>'/XY1s;'>tN\\\\\\ y >6'S$b / N h[hh k\\ h ) f kd' . 'dN O M bb 'k s h !k b 'k(,5 u o m Q , ' ', :. /14 X' d.';, ' ~,1 i 's '... A :,, l.,,,' ,;.;'.,1, X,' , / ,A. A x r. e.w w A .,, i s .~.xw. A, n is om. e.4 - A ~ - / y y / w,. h 'N6 v-y y N %., w x,4 v / 3-N 3,., g t ys y' / 3 \\ E /l\\ 7,4MLl X / \\ X4 N4 'I1MN / \\ $ XA.V'X Y)A bX / u g i yy/Awpgss< xqqsogex xggggsqu h'P,H/N NHN / /

• 2' NMYddMbDMD$<'N M M' D O M'N M3 'O

~ $l'h0Y h$'W /kk?b h ?kt 0 N ,,[ N// N's) 2 k d. 's 1,NDiOL ' x_) Ces,b ' / '.'/'slsi,'$s':f,i,',s f .-s 9 / / H v/ A. A h Vw v W A AT NA MW\\ h Ae W +s! A /A A s in T.7 (v ' / A ' / ///f/ t 'V,W (YW \\\\ W\\\\ WM /> 'Yi'rW / 9 \\ N'N)Why / h d N//// \\ X/V>Cf)AW \\ Y8A\\NM6 Mf%X / \\ XA >9'YOMM / W // X'/WXXln\\W M'dhfMX'NxW M/XAXW4NW M#M/X)W$$4/M7XYiMMW%YMM o E M86$<XM%CC(XMW'$DCM ~ ~.. . s' U - N 's-$ ^ l / O s ' 'r, i.' s l[ x'?O N hI/ E s% ~/? x' ~~ If N's l l, / N N A'r x'.N4 / A A N bA.'r eN A / A A N N A/i Y sN A / A / A6 A ( / V V VN) )(NN'V /V \\ N 'NA 'A 4NN' / V N V/i) (AY / 3 / \\ >vVK f%>f / \\ '>'d,V)' 'AN }& / \\ X/V): '/W.X / N\\x xmwy nsx xmM> xvwx xmwnMs'x x R s\\' % #WX}WKs&VAMVXWM//SWX X1N M o NM @/?OfWW 0< X1M9F'WisW @ $0W$h0 0.01 0.1

1. 0 10.O PERIOD IN SECONDS RUN 654 AUXILIARY BUILDING Figure 3A.8 Auxiliary h4WmJ envelope repse spectra calculated for spectral modulus ratios of percentiles 50, 60, 70, 80, and 90:

statistics enempass source, path, and site effects. 'Ibe statistics are inappro-priate because the criterien fr m the USGS accelerograph data is ap-plied without regard to free-field site differences. t

O o 0 N N g 3 s o. 9 k sDk /

s. N rx

,n, ,> '. : c o,2.;,. 2 5 ' M., :.. ,' a ua... : < / a \\ A x n.v. w A. x w. m.4, a .\\. a s.u m., a e, y \\ 3wx smy,- s. -c.v> x s,N - / y . s a x,,, 5 3 ,.e sh' / \\' X{AfM 't'QhX \\ X4,N['A XV fX /I\\ wx' NN 'Ji X 'QM:A' / w wm*Me v/M n usw onswx A%#Wx X%%44Wx'K NGkdMWxXHA% y MN/CO( NYdTM/IiOIMEEN',.XI NNM O I k: Sb S N m a ,,,.h N :!ru'm;t  : a w,'uw ', L'x &,.s,:n'a k w ',:,':: ; , ^,,s i m w ez u - e. , n wrma, A n.v, x i s.4 / ^ /Y Nil /Ut4 Vh5M / V N Yh ') YN'1 N' / Y N NA'v N '3 S S' / h E .'X N / '\\ /W'L W '/"W Y / \\ X'!N Y Xf1 M sX / \\ i X4 A*'8 X'1 ASX / u xsx g jfe w >< w x. yf,wxxg\\xx woenugw 6%4MW)fXWYS$keWX}%%#4WX 'k1sX\\% z NMf8.'/)OdM$$YsO('iNMib)< ')IM# ~ O $9 NthMONhCkNMbN m,r/ > xx'rn. ', :, 'x';'m.xxxc m/ . x1;;w ;;,,: / a e yo. x w m4 A m wwmv , ^ ^n wove, ^ f ///N N V /1'") X SNW / V \\\\N.'s F r'K 5"W / V \\ l' /'r'N'Y pin # / h d sX /// \\ XiN')Y'dbX / \\\\ MMS YlN X / \\ X7%') '# MAX / 7 W/%ww 'W+xxew W/n>%VM . 'NN/EWXX A% #WXIX A%#JWXl4M N'N a E MW'/DOD&%'NCODMW,'00('M/ b,,3bik hb,w-M.b,,$b[ ~ ~, \\ ~ u \\ ~... Y s' ) # a /\\ / \\ /3 h MJ V g/, / /\\ / , [\\ /?\\ [i \\ \\lt'I'IN# [\\ / /', m'h ww - A A i s w3 / Air x wor,2 v / A ^ x vw , A ( / Y N. Yh'Si X 's N 4' / V N V/ i'">Y'3N N' / 3 S. N'A ")

• N#3N'

/ 3 /1\\ 'MYX 3'Q J X / \\ X/N 'X ) M)N / \\ X/NS 't\\NsY / Nw vmxxmxN'< ven%xx wx/xx wx )1Ns'AS:Wx}'ib5 %MWXh%% o A%,SPx/x N l MM'IdOCMN'/)<)'('MD$( ld' ism ' /

• 5 bbt C 3Cn"c'Ci" O.01 0.1

1. 0 10.0 L

l PERIOD IN SECONDS RUN 689 DIESEL GENERATOR BLDG Figure 3A.9 Diesel Generator Building envelope response spectra calculated for spectral nodulus ratios of percentiles 50, 60, 70, 80, and 90: statistics enccrnpass source, path, and site effects. 'Ibe statistics are inappropriate because the criterion from the USGS accelerograph data is applied without regard to free-field site differences.

o s s s@ s o. R b b k bb b , u,h. ^y', :cucu,'x isx'ccy;w/, 'x,a 'x ;; w w,., x, o uo m/ ^ A v- / ^ A .x voxw / - an'v,%QY%v /h a ( /V \\ VAO3VY / V \\ YhO TPlN' /h M / \\ M%O3MM / \\ MWYM)N / x W.WT AIN / Nsx X'A9xxh>Nv M*xxmv ~ xewxw\\xx M A % "o 6%MM/XMA%d%/xWA% ##X/X f G'NW sM?000xW3000xW30 bbsbbkbsb}! )$MC$N g9 7x u,^x'ece= 'x,s X'c e,x w c'x, s x^c0D :t,'x-e o / n x uow - ^ x s.vrm-A ^is vsms / A ( / V \\ NWehiMY / V \\ V4DfXMN' / V \\ VfrO:WN' / h E W / \\ MMYMMX / \\ M%93CEW / l\\ M%%Y AxX / u xxx xqw,mxxx y/xqxx9ax yrxqyxvxxx A%MfvMKWs$$fM%/XMA%MZX/XM%% z xW7MMMUDOIN M 300INM 'O ~ "+ $9 NEnCa:MC R DMC::N /x,2HA 'ecNx'/, 'x,xX'x s'+w L'x / s ' ROM 0r,'x / s o / IV Pp/ x v/rWAt '. /A A N \\,%O W.A/ ^ ^ N \\NfVW /A ( / /3///\\ V/ON1NY / V \\ Vf83W/ / V \\ Vh03%I / h W M/ ///mN Mh0TAhX / \\ M%ODPAW / \\ M ZIS30XW / Nwr xrXwx9NNX X/X9xx9NNX X/X&xxvNNX MNNMh/XMh%&%/xMh4MN/XMA% o 8 M3000\\ W3000N W3000xM E-bbhC!!bbMbsf Mb!!N a ,'x /s x'c:M , x'c: ='>/x/xxc ;'Am'x ; ^' / ^ ^ x vym / ^ An vemw/ ^ ^- x vwre/ ^ '/V \\ Vfo:YWV / V \\ VA'TXVY / V \\ VhO3W/ / 'i W / \\ M%OYAM / \\ MAX X 'AM / \\ X/AI)YdW / Nsx xrXQxxspNNX X/A9xx9xNx X/X+xxvNNX-4%M&/xYh%V//MVxMk%M4/XYA% o NWM6COI@dfdY00CNM/)OIxM s S $ nCWM xC!!bbnC!!M "+ 0.01 0.1

1. 0 10.O PERIOD IN SECONDS RUN 628 SERVICE HRTER Figure 3A.10 Service Water Ptmphouse anyelope respcmse spe La calculated for spectral mcdulus ratios of percentiles 50, 60, 70, 80, and 90: statistics e. w s caurce, path, and site effects. 'the statistics are inappropriate because the critericm frcn the USGS accelerograph data is applied without regard to free-field site differences.

l

Quastion 4. Explain clearly tha relevance to your conclusions of the asser-tion on page 10 of the Addendum to Appendix B that "most dispersion is due to effects unrelated to the actual foundation / free-field phenomena under investigation." The relative response of foundation and free field to a sequence of shots is more reproducible than is indicated by the overall < apersion of the founda-tion / free-field spectral modulus ratios shown in F ares IX.D.1, 2, and 3. The fact that most of the overall dispersion is do to the scatter of free-field spectral amplitudes is evident when Fir as II.D.1, 2, and 3 are compared vich Figures VI.C.16, 20, and 28. the latter, the standard ~ deviations reflect only the source and effects. While the standard deviations differ substantially for .wo cases, the means are almost the same. Note also that varied the spectra due to random interference effects unrelated to ar __ foundation / free-field phenomena inflate the i dispersion. The implication of these observations is that the dispersiou depends on the problem addressed. If the problem is to assess the dispersion of the relative response of a foundation and an unspecified f ree-field site chosen at random, then the overall dispersion shown in Figures II.D.1, 2, and 3 is relevant. EcVever, this is not relevant if the problem is to assess for a sequence of events the reproducibility of relative r esponse of a specific site in the field, for example when ratios of foundation to accelerograph pad spectral amplitudes are computed directly, as in Figures VI.C.45 and VI.C.46. 4.1

Question 5. Given that uncertainties exist in the use of explosion tests to determine foundation response to earthquake motion, justify the use cf mean foundation / free-field spectral ratios, rather than some more conservative measure. To characterize foundation motions corresponding to the instrument pad "envel-ope spectrum

• obtained from records at Monticello Reservoir, foundation / free-field spectral ratios greater than those recommended by the Licensee should not be adopted. This follows f rom several considerations.

Firs t, "the envelope spectrum of the response spectra from data that have been recorded at Monticello is a very conservative description of ground motion Since the ground motion used in deriving the envelope was recorded on a surface soil deposit on the dam abutment, it should be consi-dered as a representation of motion at the surface." (Staf f Updated Supple-mental Testimony on Seismicity, page 44.) The assessment of the envelope as very conservative is supported by even more data through mid 1983 showing that seismicity at Monticello Reservoir has continued to decline since the above statement was written in December of 1981. Because of the conservatism used in deriving the surface ground motion envelope, it would be unduly conservative to use spectral ratios greater than mean values to derive foundation motions. This would amount to stacking conservatism upon conser-vatism. A second reason for not using foundation / free-field ratios greater than the mean values is that the values recommended by the Licensee were derived using conse rvative experimental and analytical techniques. Instruments were placed at the periphery of buildings, so that any experimental torsion induced in the buildings would be recorded as horizontal translation. It is conservative to assume that motions recorded in this manner apply to all equipment in the structure, even that at the geometrical center of the foundation.

Also, the "zero phase shift" assumption used in deriving foundation motiens is conservative relative to the use of observed phase shif t, as discussed in the 5.1

response to Question 2. Further, for frequencies below 40 Hz, computed re-sponse spectra using the "zero phase shift" filter are very close to those obtained using the unrealistically conservative " impulse filter," as discussed in the response to Question 2A. Therefore, the Licensee's recommendations are conservative. To the best of the Licensee's knowledge, there are no inherent unconservative biases or assumptions made in the design, conduct, or data analysis of this field experiment, which would reduce the effects of the censervatisms stated. The Licensee's reduced envelope spectra for the Auxiliary Building, Diesel Generator Building, and Service Water Pumphouse are conservative and justi-fied. Finally, as discussed in the respcase to Question 1, the explosion tests are appropriate to determine foundation response to earthquake motion. All observational evidence indicates that the explosion test results accurately represent the relative response of foundation and free-field sites to be expected for shallow RIS events. Therefore, the introduction of additional conservatism is not warranted. l l l l 1 5.2

/ t ',j Qu2stion 6. Waran$t the dim cbutm:nt data from Test 1, October 1981, obtained en the USGS pad? This being the case, shouldn't Figure VI.C.45 of Appendix B, rather than Figure VI.C.47, he cc pared to Figure VI.C.487 Justify or modify the estimated reduction factors in light of your answer. The dam abutment seismograph data for the October 1981 tests were not obtained on the USGS accelerograph pad. The seismometers were installed in the free field in saprolite approximately 50 feet south of the pad. Both Figures VI.C.47 and VI.C.48 show Auxiliary Building foundation / free-field spcetral modulus ratios, for Tests 4 and 1 respectively. l l l l l l l l l 6.1 .}}