WEBVTT

1
00:00:04.570 --> 00:00:06.040
Okay. Let's get started.

2
00:00:06.910 --> 00:00:13.120
Hello, everyone. My name is Jiqiu Yuan
and I am the executive director for the

3
00:00:13.450 --> 00:00:16.510
Multi-Hazard Mitigation Council, also
the Building Seismic Safety Council at

4
00:00:16.510 --> 00:00:20.770
the National Institute of Building
Sciences, known as NIBS.

5
00:00:21.160 --> 00:00:26.320
So NIBS was established by the US
Congress to ensure that all areas of

6
00:00:26.320 --> 00:00:30.960
the built environment work together to
utilize the best technology.

7
00:00:30.970 --> 00:00:35.710
Also focus on the sustainability when
creating the places where we live,

8
00:00:35.740 --> 00:00:37.900
work, learn and play.

9
00:00:38.770 --> 00:00:44.380
Developing research and convening
experts to solve problems is a part of

10
00:00:44.380 --> 00:00:45.730
our overall mission.

11
00:00:46.890 --> 00:00:51.600
The Building Seismic Safety Council,
known as BSSC, is a council with the

12
00:00:51.600 --> 00:00:55.320
NIBS established established in 1979.

13
00:00:55.470 --> 00:01:00.450
The Council's purpose is to enhance the
public safety by providing a national

14
00:01:00.450 --> 00:01:05.910
forum that fosters improved seismic
design, planning, construction and

15
00:01:05.910 --> 00:01:07.770
regulation in the building.

16
00:01:07.800 --> 00:01:09.030
In the building community.

17
00:01:09.760 --> 00:01:14.980
And we are proud to work with all our
federal also our industry partners in

18
00:01:14.980 --> 00:01:19.510
the past 40 years to help advance the
seismic safety of our nation.

19
00:01:20.140 --> 00:01:24.760
Especially, we want to thank FEMA, the
Federal Emergency Management Management

20
00:01:24.760 --> 00:01:30.280
Agency, for their support to produce
the NEHRP recommended seismic

21
00:01:30.280 --> 00:01:35.950
provisions. Also, the related
supporting documents, including today's

22
00:01:36.790 --> 00:01:40.330
2020 NEHRP Provisions Design Example
series.

23
00:01:41.080 --> 00:01:46.840
So today is our eighth webinar in our
NEHRP webinar series.

24
00:01:47.350 --> 00:01:51.760
In early ones, we have covered topics,
many different topics that were

25
00:01:51.760 --> 00:01:58.000
introduced by the 2020 NEHRP and NEHRP
provisions like the diaphragm design,

26
00:01:58.150 --> 00:02:03.370
carport, concrete shear walls the new
non-structural component design

27
00:02:03.370 --> 00:02:08.320
equations. So if you missed any of
them, you can listen to the recording

28
00:02:08.320 --> 00:02:10.150
on the BSSC website.

29
00:02:11.140 --> 00:02:16.720
And today, it's our great pleasure to
have a Ron and Sanaz with us today to

30
00:02:16.720 --> 00:02:19.400
talk about the evolution of the seismic
design values.

31
00:02:19.420 --> 00:02:22.960
Also the USGS national seismic hazard
model.

32
00:02:23.910 --> 00:02:27.890
So Ron is a senior principal with SDH.

33
00:02:28.830 --> 00:02:33.300
He has nearly 50 years of a structure
in engineering, design, education,

34
00:02:33.300 --> 00:02:36.600
construction and theory investment
experience.

35
00:02:37.230 --> 00:02:43.470
Currently currently completing a second
term as the chair of the ASCE 7

36
00:02:43.500 --> 00:02:44.970
Standards Committee.

37
00:02:45.270 --> 00:02:49.760
He has been an active participant in
the NEHRP provisions in the NEHRP

38
00:02:50.160 --> 00:02:53.670
provisions update process since 1996.

39
00:02:54.060 --> 00:03:00.570
And there was a member of the Joint
USGS and the BSSC panel that developed

40
00:03:00.660 --> 00:03:06.690
the basis for the motion maps adopted
by the provisions in 1997 and

41
00:03:07.350 --> 00:03:11.520
2007 and the chaired the panel in 2017.

42
00:03:12.440 --> 00:03:19.340
He served as a chair for the 2020, 2003
and 2009 cycle, and the

43
00:03:19.610 --> 00:03:23.660
recognition of his work in developing
in developing and improving the

44
00:03:23.660 --> 00:03:25.190
nation's building codes.

45
00:03:25.550 --> 00:03:32.300
He has been awarded the NCSEA's James
Delehay Award, SEI's

46
00:03:32.300 --> 00:03:38.750
Walter P. Moore Award and the Newmark
Medal and was elected to the to the

47
00:03:38.750 --> 00:03:41.420
National Academy of Engineers in 2015.

48
00:03:43.950 --> 00:03:46.980
Our second speaker. Speaker today is
Sanaz Rezaeian.

49
00:03:48.090 --> 00:03:53.970
She is a research structural engineer
at the US Geological Survey, USGS.

50
00:03:54.750 --> 00:03:59.550
She graduated from the UC Berkeley in
2010 and joined the USGS in 2011.

51
00:04:00.670 --> 00:04:06.870
Sanaz specializes in earthquake
engineering, stochastic simulation of

52
00:04:06.870 --> 00:04:11.100
grand motions and the modeling of grand
motion and prediction earthquakes

53
00:04:11.220 --> 00:04:16.530
equations. She has been a collaborator
with the next generation attenuation

54
00:04:16.530 --> 00:04:22.680
known as MTA Project since 2010 and the
co-leader of the Southern California

55
00:04:22.680 --> 00:04:28.230
Earthquake Center's the Motion
Simulation Validation Group since 2012

56
00:04:28.470 --> 00:04:33.390
and also a USGS liaison on the Building
Seismic Safety Council's Seismic Safety

57
00:04:33.390 --> 00:04:35.250
Council since 2015.

58
00:04:35.550 --> 00:04:39.150
Again, great pleasure to have us to
have put you on stage here today.

59
00:04:39.450 --> 00:04:44.760
Before I pass this to Ron to start the
presentation, let me do a few quick

60
00:04:44.910 --> 00:04:46.350
housekeeping items here.

61
00:04:46.770 --> 00:04:51.390
So all the participants will be able to
type in questions through the Q&A box,

62
00:04:51.810 --> 00:04:56.400
and we strongly encourage you to do
that during the presentation, and we

63
00:04:56.400 --> 00:05:00.300
will try to answer most of them in the
end of the presentation.

64
00:05:01.350 --> 00:05:03.870
Second, the session will be recorded.

65
00:05:04.170 --> 00:05:08.160
The recording will be available on the
NIBS's BSSC website.

66
00:05:08.790 --> 00:05:11.410
Also, we will provide the PDFs.

67
00:05:11.490 --> 00:05:17.580
Also the C use the continue education
unit for AI and ICC and participants.

68
00:05:18.270 --> 00:05:24.540
So for those who attend live webinar,
webinar, webinar today, so look for our

69
00:05:24.540 --> 00:05:29.670
follow up emails for the both the
recording, also the the certificates.

70
00:05:30.000 --> 00:05:32.850
So without further ado, our past days
to Ron.

71
00:05:33.150 --> 00:05:33.720
Thank you.

72
00:05:35.000 --> 00:05:36.080
Thank you, Jiqiu.

73
00:05:44.400 --> 00:05:48.280
Hello, everyone. I'm very happy to be
here with you today.

74
00:05:49.920 --> 00:05:56.790
I'm going to be giving you one of my
favorite talks on seismic design,

75
00:05:56.970 --> 00:06:02.010
which is the evolution of seismic
design over the period of the last

76
00:06:02.030 --> 00:06:03.450
hundred years or so.

77
00:06:04.350 --> 00:06:10.200
In particular, I'm going to be looking
at how the maps and the effect of the

78
00:06:10.200 --> 00:06:13.050
maps on the design values we've used.

79
00:06:13.050 --> 00:06:18.090
Basically, the base shear forces we use
in designing structures have evolved

80
00:06:18.090 --> 00:06:23.030
over time. And I'm going to be doing
that basically as a setup for Sanaz,

81
00:06:23.030 --> 00:06:25.190
who will be following me.

82
00:06:25.190 --> 00:06:28.350
I'm going to give you a broad overview
of how we've gotten to where we are

83
00:06:28.380 --> 00:06:33.750
today. And then Sanaz we'll give you a
detailed dive into exactly what is

84
00:06:33.750 --> 00:06:40.200
behind the maps that appear in the
current NEHRP provisions and also ASCE

85
00:06:40.230 --> 00:06:45.810
7-22 and eventually the International
Building Code 2024 edition.

86
00:06:51.440 --> 00:06:54.950
Since the earliest inception of the
building codes.

87
00:06:57.210 --> 00:07:02.160
The design forces that we've used this
proportion, our structures have been

88
00:07:02.160 --> 00:07:04.320
based on seismic risk maps.

89
00:07:04.440 --> 00:07:09.660
Over in the upper left hand corner is a
map that I took from the 1935 Uniform

90
00:07:09.660 --> 00:07:13.590
Building Code was called the Pacific
Coast Building Code in those days.

91
00:07:14.400 --> 00:07:20.040
And down on the lower left hand corner,
a series of maps from a recent division

92
00:07:20.040 --> 00:07:23.100
of the NEHRP provisions, not the latest
edition.

93
00:07:24.830 --> 00:07:27.680
The maps are based on science.

94
00:07:29.060 --> 00:07:33.410
But what we do as engineers and
designing buildings is not science.

95
00:07:33.710 --> 00:07:38.930
Rather, it's an application of
engineering judgment that's informed by

96
00:07:38.930 --> 00:07:44.090
science. And when the maps are
developed and adopted into the

97
00:07:44.090 --> 00:07:49.130
provisions or the building code, there
is a balancing act between science on

98
00:07:49.130 --> 00:07:54.290
the one hand and what it tells us and
engineering judgment on the other.

99
00:07:54.620 --> 00:08:00.290
And finally, not to some small extent,
also politics and the willingness of

100
00:08:00.290 --> 00:08:03.740
society to design for certain things or
not.

101
00:08:04.130 --> 00:08:07.460
So I will be telling you about that
story.

102
00:08:08.840 --> 00:08:15.080
Since 2000 or so, seismic design maps
have resulted from collaboration of the

103
00:08:15.080 --> 00:08:19.880
USGS, who has handled the Science and
the Building Seismic Safety Council

104
00:08:19.880 --> 00:08:22.310
under NIBS, who's handled the
engineering.

105
00:08:22.310 --> 00:08:26.810
And this collaboration has resulted in
the maps that we have today.

106
00:08:29.250 --> 00:08:30.750
Let's start at the beginning.

107
00:08:30.870 --> 00:08:32.490
Always a good place to start.

108
00:08:34.020 --> 00:08:37.860
What's generally acknowledged in the
United States as being the first

109
00:08:37.860 --> 00:08:43.470
building code that had seismic design
provisions was the 1927 edition of the

110
00:08:43.470 --> 00:08:46.620
Uniform Building Code shown here on the
left.

111
00:08:47.460 --> 00:08:53.070
In fact, that was not the first seismic
design code either in the United States

112
00:08:53.070 --> 00:08:58.860
or internationally following the 1868
earthquake that occurred near Hayward,

113
00:08:58.860 --> 00:09:04.320
California. Some communities in the San
Francisco Bay area actually started to

114
00:09:04.320 --> 00:09:08.850
adopt seismic design requirements in
their codes, although they were not

115
00:09:08.850 --> 00:09:11.040
specifically identified as such.

116
00:09:12.200 --> 00:09:15.320
And around 1910 or so.

117
00:09:16.260 --> 00:09:18.900
Italy, Japan and the United States.

118
00:09:19.860 --> 00:09:25.260
Independently began to develop seismic
design requirements and their building

119
00:09:25.260 --> 00:09:32.070
codes. Anyhow, looking at the 1927
Uniform Building Code, it actually did

120
00:09:32.070 --> 00:09:34.350
not have a seismic map in it.

121
00:09:34.740 --> 00:09:38.740
And in fact, the seismic provisions
were not really in the code either.

122
00:09:38.790 --> 00:09:42.090
They were in a non-mandatory appendix.

123
00:09:43.060 --> 00:09:48.430
And basically the philosophy in the
1927 Uniform Building Code was that if

124
00:09:48.430 --> 00:09:52.510
you wanted to design for earthquake
resistance, then you designed for it.

125
00:09:52.630 --> 00:09:56.400
You didn't need a map to tell you how
much design you needed.

126
00:09:56.410 --> 00:10:00.190
If you were going to design for it, you
went to The Full Monty and you designed

127
00:10:00.190 --> 00:10:04.050
for it. And if you didn't want to
design for earthquake, you didn't

128
00:10:04.050 --> 00:10:10.320
design for it. Basically, the 1927
Uniform Building Code had two

129
00:10:10.320 --> 00:10:12.900
independent criteria for seismic
design.

130
00:10:13.500 --> 00:10:19.350
If you are on soft soils, soils with an
allowable load bearing pressure of two

131
00:10:19.350 --> 00:10:20.730
tons per square foot.

132
00:10:22.410 --> 00:10:27.930
Or less, then you would design for 10%
of the weight of the structure applied

133
00:10:27.930 --> 00:10:29.400
as a lateral force.

134
00:10:29.820 --> 00:10:34.290
And if you were on competent soil soils
that had two tons per square foot or

135
00:10:34.290 --> 00:10:38.700
greater bearing pressure, then you got
a slight discount on that.

136
00:10:39.810 --> 00:10:43.920
You were able to design for seven and a
half percent of the weight of the

137
00:10:43.920 --> 00:10:46.470
building applied as a lateral force at
each level.

138
00:10:47.160 --> 00:10:52.050
The reason for the reduction in the
allowable lateral force requirements

139
00:10:52.050 --> 00:10:57.990
for buildings on firm soils was that in
the 1906 San Francisco earthquake, it

140
00:10:57.990 --> 00:11:02.430
had been recognized by engineers that
buildings that were located on soft

141
00:11:02.430 --> 00:11:07.320
soils performed much more poorly than
buildings that were located on firmer

142
00:11:07.320 --> 00:11:13.560
soils. So this lesson made it into the
1927 Uniform Building Code, but it was

143
00:11:13.560 --> 00:11:15.570
forgotten a few years later.

144
00:11:16.910 --> 00:11:20.780
The other thing that's interesting
about the 10% W criteria use the

145
00:11:20.780 --> 00:11:26.750
buildings on most soils was that the
same criteria was adopted both in Japan

146
00:11:26.750 --> 00:11:30.410
and in Italy at about the same point in
time.

147
00:11:30.770 --> 00:11:35.330
I don't know if the three countries
engineers and scientists communicated

148
00:11:35.330 --> 00:11:37.640
with each other back in the 1920s.

149
00:11:37.640 --> 00:11:41.090
But it is interesting that all three
countries adopted the lateral course

150
00:11:41.090 --> 00:11:46.130
design criteria of equivalent lateral
force is equal to 10% of the building

151
00:11:46.130 --> 00:11:49.400
weight applied uniformly at each level
of the structure.

152
00:11:52.310 --> 00:11:57.260
The first seismic risk map appearing in
US building codes actually occurred in

153
00:11:57.260 --> 00:12:00.620
the 1935 edition of the Uniform
Building Code.

154
00:12:01.460 --> 00:12:06.980
It covered the Western United States,
only shown over here on the left side

155
00:12:06.980 --> 00:12:11.600
of the slide, and it provided
qualitative write ratings of seismic

156
00:12:11.600 --> 00:12:15.740
risk based on the observed seismicity
in the region.

157
00:12:16.310 --> 00:12:22.220
And basically, it had three zones, zone
one, which was where earthquakes don't

158
00:12:22.220 --> 00:12:23.240
really occur.

159
00:12:24.020 --> 00:12:28.070
Zone three, where really big
earthquakes have occurred like the 1906

160
00:12:28.070 --> 00:12:32.960
San Francisco earthquake and zone, two
areas that have experienced shaking in

161
00:12:32.960 --> 00:12:38.240
historic times but have not been really
close to really destructive

162
00:12:38.240 --> 00:12:44.070
earthquakes. The design values in the
code were not directly tied to

163
00:12:44.070 --> 00:12:45.950
anticipated ground acceleration.

164
00:12:45.960 --> 00:12:50.820
In fact, I'm not even sure that the
framers of the 1935 Uniform Building

165
00:12:50.820 --> 00:12:55.260
Code had a good understanding of just
how strong the ground could shake.

166
00:12:56.320 --> 00:13:00.790
They expressed ground shaking intensity
and sort of three forms really

167
00:13:00.790 --> 00:13:05.290
destructive. Never happens or happens
but isn't too bad.

168
00:13:06.800 --> 00:13:11.990
They kept the design criteria very
similar to what was in the earlier

169
00:13:11.990 --> 00:13:18.550
building code. If you were in Zone
three

170
00:13:18.910 --> 00:13:24.310
and you are on firm soils, you designed
for 8% of the building weight as a

171
00:13:24.310 --> 00:13:29.890
lateral force. And if you are on soft
soils, 16% of the building's weight.

172
00:13:30.040 --> 00:13:34.480
And this begins with only the second
edition of the Uniform Building Code of

173
00:13:34.480 --> 00:13:39.010
an effect that I will refer to as a
pogo stick, where the required seismic

174
00:13:39.010 --> 00:13:45.610
design values jump up and down in time
as additions of the code change and

175
00:13:45.610 --> 00:13:50.920
responds to engineers and scientists
understanding of what is necessary to

176
00:13:50.920 --> 00:13:54.850
provide us with buildings that will
perform acceptably.

177
00:13:54.850 --> 00:13:56.010
And it is, I.

178
00:13:56.410 --> 00:14:01.480
I will admit, quite irritating to me as
a practitioner and I'm sure to many of

179
00:14:01.480 --> 00:14:04.930
you as well, when these design values
go up and down from addition of the

180
00:14:04.930 --> 00:14:09.220
code to additions of the code, there
are always good reasons for it.

181
00:14:09.850 --> 00:14:15.100
We have tried in recent times to
prevent it, but we have found that the

182
00:14:15.100 --> 00:14:17.800
good reasons outweigh the reasons not
to.

183
00:14:17.830 --> 00:14:20.890
And so the values do jump up and down
from code.

184
00:14:20.890 --> 00:14:27.760
The code. What you will note here is
that in the 1935 UBC, if you

185
00:14:27.760 --> 00:14:33.610
were in zone two, you designed for half
the force required in Zone three.

186
00:14:33.700 --> 00:14:38.260
And if you were in Zone one, you
designed for a quarter of the force

187
00:14:38.440 --> 00:14:40.150
that was required in Zone three.

188
00:14:40.150 --> 00:14:45.370
And these relations remained in place
for many, many years, really right up

189
00:14:45.370 --> 00:14:51.610
until the 1990s when we abandoned zones
and went to seismic design.

190
00:14:51.630 --> 00:14:52.930
Again, the contours.

191
00:14:56.860 --> 00:15:03.190
In the 1949 edition of the UBC, the
maps went national for

192
00:15:03.190 --> 00:15:09.670
the first time and for reference
purposes, the seismic

193
00:15:09.750 --> 00:15:15.340
zonation map actually showed intensity
dots showing the locations of historic

194
00:15:15.340 --> 00:15:20.500
earthquakes in the United States with
small dots representing low intensity

195
00:15:20.500 --> 00:15:24.980
earthquakes and large dots representing
high intensity earthquakes.

196
00:15:25.000 --> 00:15:30.490
And you can see the direct tie between
historic seismicity in our region and

197
00:15:30.490 --> 00:15:32.830
the required seismic design.

198
00:15:34.080 --> 00:15:37.080
The 1949 edition of the UBC.

199
00:15:38.560 --> 00:15:43.660
Identified four seismic zones zone
zero, zone one, zone two, zone three.

200
00:15:44.620 --> 00:15:49.660
Zone one was identified as regions
where there was no historic damage from

201
00:15:49.660 --> 00:15:54.610
earthquakes. Zone one regions where
there was minor damage.

202
00:15:56.180 --> 00:15:58.250
Zone two, moderate damage.

203
00:15:58.490 --> 00:16:00.440
Zone three, major damage.

204
00:16:01.390 --> 00:16:05.620
And then there was a base year formula
that was different from that contained

205
00:16:05.620 --> 00:16:11.920
in the 1935 UBC, where base year was
tied loosely to spectral response.

206
00:16:11.920 --> 00:16:16.390
And when I mean loosely, I mean as
buildings got taller, the required

207
00:16:16.390 --> 00:16:18.730
seismic design forces reduced.

208
00:16:19.490 --> 00:16:21.860
And again talking about the pogo stick.

209
00:16:23.000 --> 00:16:26.990
Whereas in the last edition of the
code, the required design force in

210
00:16:26.990 --> 00:16:30.800
seismic zone three on soft soils was.

211
00:16:32.860 --> 00:16:38.650
8%. In this edition of the code for
single story buildings in Zone three,

212
00:16:38.740 --> 00:16:41.290
it jumped up to 13%.

213
00:16:41.740 --> 00:16:46.270
And by the way, one of the things that
got forgotten in 1949 was the

214
00:16:46.270 --> 00:16:52.060
importance of site class effects, the
stiffness and depth of soils at a site

215
00:16:52.060 --> 00:16:54.070
on the intensity of ground shaking.

216
00:16:54.100 --> 00:16:58.180
That effect disappeared completely for
the code and stayed out of the code for

217
00:16:58.180 --> 00:16:59.590
the next 30 years.

218
00:17:02.720 --> 00:17:08.360
In the 1960s, the map looked very
similar to what was in the 1949 edition

219
00:17:08.360 --> 00:17:09.380
of the UBC.

220
00:17:09.950 --> 00:17:14.240
But now there was a clear tie and a
legend in the lower left hand corner of

221
00:17:14.240 --> 00:17:19.790
the map to intensity in the form of
modified mortality intensity.

222
00:17:20.210 --> 00:17:24.890
Modified mortality intensity as a
scale, a qualitative scale that

223
00:17:24.890 --> 00:17:29.360
identifies the intensity of ground
shaking based on the typical effects of

224
00:17:29.360 --> 00:17:31.010
buildings of different types.

225
00:17:31.680 --> 00:17:38.340
In the 1960s and through 1973, there's
a legend on the seismic donation map

226
00:17:38.340 --> 00:17:43.470
that says Zone One relates to regions
where only minor damage is expecting

227
00:17:43.470 --> 00:17:46.800
corresponding to earthquake ground
shaking.

228
00:17:46.800 --> 00:17:50.460
Of mmi intensity five or six.

229
00:17:50.460 --> 00:17:55.500
Zone two regions of moderate damage
potential where mmi ground shaking of

230
00:17:55.500 --> 00:18:02.490
intensity seven could be expected and
zone three regions where mmi

231
00:18:02.520 --> 00:18:06.210
eight or higher intensity ground motion
could be expected.

232
00:18:07.710 --> 00:18:12.840
The link to spectral acceleration
values was improved in the base year

233
00:18:12.870 --> 00:18:19.770
formula in the 1960s and early 1970s,
where now the base year requirement is

234
00:18:19.770 --> 00:18:25.590
clearly tied to the period of the
structure which is now denoted by T.

235
00:18:26.840 --> 00:18:29.060
And the pogo stick comes back down
again.

236
00:18:29.060 --> 00:18:34.700
And we're reducing from a 13% design
base sheer force for the most

237
00:18:34.700 --> 00:18:36.950
susceptible buildings to 10%.

238
00:18:37.910 --> 00:18:39.110
In the seismic zones.

239
00:18:39.110 --> 00:18:40.490
We still have a relationship.

240
00:18:40.490 --> 00:18:45.970
That Zone two is designed for half the
force of zone three and Zone one.

241
00:18:45.980 --> 00:18:48.740
A quarter of the force of zone three.

242
00:18:53.420 --> 00:18:57.830
Major strides were made in seismic
resistant design requirements in the

243
00:18:57.830 --> 00:18:59.870
1976 UBC.

244
00:19:00.590 --> 00:19:06.890
This was largely as a result of lessons
that were learned in the 1971 San

245
00:19:06.890 --> 00:19:10.580
Fernando earthquake that occurred in
the Los Angeles, California area.

246
00:19:11.480 --> 00:19:16.280
Prior to the 71 San Fernando
earthquake, California structural

247
00:19:16.280 --> 00:19:20.570
engineers felt that they really knew
how to design for earthquakes as well,

248
00:19:20.570 --> 00:19:24.170
and the building code did everything
that it needed to do.

249
00:19:24.930 --> 00:19:30.870
Well, in February of 1971, there was a
moderate magnitude earthquake,

250
00:19:30.870 --> 00:19:36.150
magnitude 6.8, that occurred in the San
Fernando Valley, north of Los Angeles,

251
00:19:36.150 --> 00:19:40.830
and created a great deal of damage to
modern construction designed to

252
00:19:40.830 --> 00:19:42.540
contemporary building codes.

253
00:19:42.660 --> 00:19:46.650
And this caused engineers to rethink
things quite a lot.

254
00:19:47.490 --> 00:19:53.550
And one of the things they did is they
reintroduced soil site class effects

255
00:19:54.030 --> 00:19:58.500
and adjusted the seismic design force
based on the site class that was

256
00:19:58.500 --> 00:20:00.060
present at the site.

257
00:20:01.250 --> 00:20:06.590
We also bumped up the required base 
shear again, the pogo stick acting once

258
00:20:06.590 --> 00:20:10.430
again from 10% back to 14%.

259
00:20:11.800 --> 00:20:15.940
We also introduced an occupancy
importance factor, recognizing that we

260
00:20:15.940 --> 00:20:20.950
want better performance out of some
types of structures, and finally

261
00:20:20.950 --> 00:20:27.430
introduced a new seismic zone zone four
limited to those portions of

262
00:20:27.430 --> 00:20:31.860
California that were close to major
active faults.

263
00:20:31.870 --> 00:20:34.810
And the sort of looking at the lower
left hand corner here.

264
00:20:35.140 --> 00:20:39.130
The strip that you see running north to
south along the coast of California

265
00:20:39.790 --> 00:20:42.220
represents the San Andreas fault.

266
00:20:42.670 --> 00:20:49.180
Then some faults that extend around to
California's Central Valley, the San

267
00:20:49.210 --> 00:20:53.620
Jacinto fault, and then the Owens
Valley fault running up into Nevada.

268
00:20:54.280 --> 00:20:59.920
These were faults that were by 1976,
recognized as capable of very large

269
00:20:59.920 --> 00:21:03.220
magnitude earthquakes and very
destructive ground shaking.

270
00:21:03.850 --> 00:21:08.290
And so extra force requirements were
instituted near these faults.

271
00:21:12.190 --> 00:21:18.190
In addition to resulting in major
changes in the UBC in the 1976 edition,

272
00:21:18.190 --> 00:21:23.320
the 71 San Fernando earthquake also
spurred a very important project

273
00:21:23.320 --> 00:21:27.100
undertaken by the Applied Technology
Council, which ended with the

274
00:21:27.100 --> 00:21:32.590
publication in 1978 of the ATC-03
Report.

275
00:21:34.230 --> 00:21:40.920
ATC-03 took a fresh look at the tie
between seismic design forces

276
00:21:40.920 --> 00:21:46.410
and ground motion and introduced
separate zonation maps depending on

277
00:21:46.410 --> 00:21:49.530
whether you had a short period or a
long period structure.

278
00:21:50.550 --> 00:21:57.390
The first map was a map of effective
peak ground accelerations known as

279
00:21:57.390 --> 00:22:03.870
A sub a. And the second map, A sub v
showed contours of effective peak

280
00:22:03.870 --> 00:22:06.180
velocity related acceleration.

281
00:22:06.600 --> 00:22:10.900
Later on, our building codes would
begin to call these values S sub s and

282
00:22:11.040 --> 00:22:15.330
S sub b one. But these were first
introduced in ATC-03.

283
00:22:15.960 --> 00:22:22.770
In the 1978 edition ATC-03 maintain the
concept of seismic zones,

284
00:22:22.770 --> 00:22:26.940
although it showed contours of
effective peak acceleration and

285
00:22:26.940 --> 00:22:28.470
effective peak velocity.

286
00:22:29.220 --> 00:22:34.620
It also established seismic zones with
uniform values of these coefficients

287
00:22:34.620 --> 00:22:35.850
throughout the zone.

288
00:22:37.690 --> 00:22:43.480
And declared that for the first time
that the values of A sub a and A sub v

289
00:22:43.480 --> 00:22:48.610
that were mapped in these seismic
hazard maps corresponded to ground

290
00:22:48.610 --> 00:22:53.830
motion that approximated the level of
shaking that would have a 10% 50 year

291
00:22:53.830 --> 00:22:58.900
exceedance probability or a return
period of 475 years.

292
00:22:59.820 --> 00:23:06.360
And for the next 20 years or so, 475
year ground motion would remain as the

293
00:23:06.360 --> 00:23:09.930
basis for seismic design in US building
codes.

294
00:23:10.710 --> 00:23:17.250
In ATC-03, the Bashir equations were
directly, directly tied to A sub a and

295
00:23:17.250 --> 00:23:23.040
A sub v and clearly identified as being
spectral response parameters.

296
00:23:26.520 --> 00:23:32.970
The 1988 Uniform Building Code
underwent a major revision of its

297
00:23:32.970 --> 00:23:39.270
seismic design criteria, and mostly it
adopted the requirements of ATSC 306.

298
00:23:40.360 --> 00:23:44.740
It did retain a single seismic zone map
which was constructed based on the

299
00:23:44.750 --> 00:23:51.370
asthma and ACB maps contained in 83 Z,
and it used the

300
00:23:51.370 --> 00:23:58.210
parameter Z that was still called
zonation parameter, but was really now

301
00:23:58.210 --> 00:24:03.970
the effect of peak ground acceleration
A sub a the base year equation for the

302
00:24:03.970 --> 00:24:09.040
first time was reformulated in a
building code to the format that we

303
00:24:09.040 --> 00:24:14.020
recognize today as being a function of
spectral response and the weight of the

304
00:24:14.020 --> 00:24:17.650
building. And the pogo stick jumped
once again.

305
00:24:17.770 --> 00:24:22.870
And we're now designing for a worst
case based share of almost 20% of the

306
00:24:22.870 --> 00:24:23.890
weight of the building.

307
00:24:27.190 --> 00:24:33.850
After publication of ATC three in 1978,
the Federal Emergency Management

308
00:24:33.850 --> 00:24:38.800
Agency began to fund to fund the
Building Seismic Safety Council to

309
00:24:38.800 --> 00:24:40.840
develop the NEHRP provisions.

310
00:24:41.560 --> 00:24:45.760
The NEHRP provisions basically were
based on the ATC three document.

311
00:24:46.660 --> 00:24:52.060
BSSC convened a panel of volunteers at
a call of the Provisions Update

312
00:24:52.060 --> 00:24:55.270
Committee. We were on a periodic basis
originally.

313
00:24:55.270 --> 00:25:01.500
Three years would review the ATC three
document, make changes to it that they

314
00:25:01.510 --> 00:25:05.980
were deemed necessary to make them
practical and available for adoption as

315
00:25:05.980 --> 00:25:09.010
a building code and publish them as the
NEHRP provisions.

316
00:25:10.450 --> 00:25:15.580
And in the early 1990s, both the
Southern Standard Building Code,

317
00:25:15.580 --> 00:25:21.730
Congress International and the building
officials and code administrators who

318
00:25:21.730 --> 00:25:26.110
published the Standard Building Code in
the South and the National Building

319
00:25:26.110 --> 00:25:30.880
Code in the Northeast adopted their
seismic design criteria by basically

320
00:25:30.880 --> 00:25:35.860
transcribing the 1988 edition of the 
NEHRP provisions into their code.

321
00:25:36.100 --> 00:25:40.420
This was the first formal adoption of
the NEHRP provisions by building codes,

322
00:25:41.380 --> 00:25:46.630
and this adoption they retained 500
year seismic risk maps with a sub a and

323
00:25:46.630 --> 00:25:53.350
a sub v contours, and they retain the
strength level base shear forces, which

324
00:25:53.350 --> 00:25:59.020
were basically 140% of the design
forces contained in the contemporary

325
00:25:59.020 --> 00:26:00.340
uniform building code.

326
00:26:02.300 --> 00:26:09.050
Ultimately in 1997, fearing that they
would be viewed as being left behind by

327
00:26:09.050 --> 00:26:11.900
the other two building code
organizations.

328
00:26:11.900 --> 00:26:15.170
The International Conference of
Building Officials would publish.

329
00:26:15.170 --> 00:26:20.750
The Uniform Building Code also adopted
strength level forces in their building

330
00:26:20.750 --> 00:26:26.960
code and revised their design maps to
have C, sub, A and C sub

331
00:26:26.960 --> 00:26:32.510
values similar to the A, sub A and ace
of the values contained in the NEHRP

332
00:26:32.510 --> 00:26:37.580
provisions and very similar to the se
of the sense of d one values that we

333
00:26:37.580 --> 00:26:39.050
design for today.

334
00:26:40.690 --> 00:26:46.510
By the mid 1990s, the three building
code development organizations SBS,

335
00:26:46.510 --> 00:26:52.930
CCI, Bowker and ICBO had decided that
they were going to collaborate and set.

336
00:26:53.290 --> 00:26:58.300
The year 2000 is a date when instead of
publishing three individual building

337
00:26:58.300 --> 00:27:03.160
codes, they would publish one single
building code that we now call the

338
00:27:03.160 --> 00:27:04.900
International Building Code.

339
00:27:06.460 --> 00:27:13.420
This caused a deal of discussion in
engineering and seismology levels as to

340
00:27:13.420 --> 00:27:19.780
how we could best provide a really
suitable series of maps that could be

341
00:27:19.780 --> 00:27:23.050
used for seismic design in the United
States.

342
00:27:25.420 --> 00:27:31.390
In response to this, BSSC, together
with USGS,

343
00:27:31.870 --> 00:27:37.240
convened a panel of experts consisting
of volunteers from around the United

344
00:27:37.240 --> 00:27:40.630
States structural engineers,
geotechnical engineers and

345
00:27:40.630 --> 00:27:46.360
seismologists to see how they could go
about crafting a series of nationally

346
00:27:46.360 --> 00:27:49.720
applicable maps for use in seismic
design.

347
00:27:51.920 --> 00:27:55.730
It was very quickly decided by the
panel working on this that the

348
00:27:55.730 --> 00:28:00.650
seismicity in the eastern US was very
different from the seismicity in the

349
00:28:00.650 --> 00:28:01.790
western US.

350
00:28:02.150 --> 00:28:06.920
Very large earthquakes they now knew
could occur in the eastern US, even

351
00:28:06.920 --> 00:28:09.560
though they had not done so in historic
times.

352
00:28:11.330 --> 00:28:13.280
But they do so very rarely.

353
00:28:13.280 --> 00:28:16.700
Whereas in the western US large
magnitude, earthquakes occur

354
00:28:16.700 --> 00:28:23.440
frequently. It was recognized that the
500 year return period used for

355
00:28:23.440 --> 00:28:27.130
ground motion maps and the then NEHRP 
provisions.

356
00:28:28.040 --> 00:28:32.540
And in the building codes of the time
were not really adequate to capture the

357
00:28:32.540 --> 00:28:37.070
levels of ground shaking that would
occur from major earthquakes that had

358
00:28:37.070 --> 00:28:42.860
occurred in the eastern US, including
the 1811 1812 New Madrid series and the

359
00:28:42.860 --> 00:28:46.310
large earthquake that occurred near
Charleston, South Carolina, in the

360
00:28:46.310 --> 00:28:53.000
1890s. And so there was a push to start
adopting ground motion

361
00:28:53.000 --> 00:28:56.690
maps that would have longer return
periods.

362
00:28:57.840 --> 00:29:02.760
In the Western US, these longer return
periods would result in ground motion

363
00:29:02.760 --> 00:29:09.090
values that just did not seem realistic
or practical for design for engineers

364
00:29:09.090 --> 00:29:10.770
in the western United States.

365
00:29:11.540 --> 00:29:18.290
And the Seismic Design Values Panel put
together by DSC and USGS basically was

366
00:29:18.290 --> 00:29:23.480
unsuccessful. It could not reach a
compromise that would satisfy both the

367
00:29:23.480 --> 00:29:27.650
needs to capture the large ground
motions in the eastern US that could

368
00:29:27.650 --> 00:29:33.260
occur and what Western engineers viewed
as practical for design.

369
00:29:38.460 --> 00:29:45.210
By the time this was occurring, the
USGS was developing maps using a

370
00:29:45.210 --> 00:29:48.630
approach known as probabilistic seismic
hazard analysis.

371
00:29:48.960 --> 00:29:52.080
And I'll do a very brief introduction
of how this is done.

372
00:29:52.820 --> 00:29:56.690
Let's say you want to determine what
the probability of ground motion at a

373
00:29:56.690 --> 00:29:57.890
specific site is.

374
00:29:57.890 --> 00:30:02.570
And for this little example here, I'm
using the site in the Central Valley of

375
00:30:02.570 --> 00:30:05.660
California, shown by a star on the map.

376
00:30:07.170 --> 00:30:11.550
You start by looking at all of the
known sources or faults that could

377
00:30:11.550 --> 00:30:13.860
generate strong ground motion at the
site.

378
00:30:14.100 --> 00:30:16.980
And we'll pick this one little fault
here.

379
00:30:17.370 --> 00:30:19.320
Over in eastern California.

380
00:30:21.150 --> 00:30:25.890
We want to identify what the distance
of that site from that particular

381
00:30:25.890 --> 00:30:27.120
source is.

382
00:30:29.820 --> 00:30:33.990
Then we want to look at magnitude
recurrence relationships for that

383
00:30:33.990 --> 00:30:38.760
particular fault and understand what
the probability of experiencing

384
00:30:39.180 --> 00:30:42.360
earthquakes of different magnitude on
that fault are.

385
00:30:44.660 --> 00:30:49.280
We'll pick a given probability of
exceedance and find out what the

386
00:30:49.280 --> 00:30:52.760
magnitude of earthquake likely to occur
on that fault is.

387
00:30:52.760 --> 00:30:57.590
In this case, I've taken a probability
of exceedance of 0.07 per year and

388
00:30:57.590 --> 00:31:01.070
found that that fault at that
probability level is capable of

389
00:31:01.070 --> 00:31:03.230
generating a magnitude six earthquake.

390
00:31:05.010 --> 00:31:09.750
We then go into what used to be called
attenuation relationships and more

391
00:31:09.750 --> 00:31:14.160
recently have been called ground ground
motion prediction equations or ground

392
00:31:14.160 --> 00:31:19.890
motion prediction models, which tell us
for a given series of parameters,

393
00:31:19.890 --> 00:31:23.940
including earthquake magnitude,
distance of the fault from the site

394
00:31:23.970 --> 00:31:26.490
type of soil that is present at the
site.

395
00:31:26.580 --> 00:31:31.560
What the probabilistic distribution of
ground motion intensity is at that

396
00:31:31.560 --> 00:31:36.540
site. And we find that for a magnitude
six earthquake, we have a mean ground

397
00:31:36.540 --> 00:31:42.060
motion at this site located at a
distance of 20% of G.

398
00:31:43.300 --> 00:31:49.090
We do this for all different
probabilities on that fault and for all

399
00:31:49.090 --> 00:31:51.970
of the other faults that are
surrounding the site.

400
00:31:52.150 --> 00:31:56.740
And we add up the results of that, and
it allows us to construct a probability

401
00:31:56.740 --> 00:32:02.410
distribution that tells us at different
levels of probability what the maximum

402
00:32:02.410 --> 00:32:04.810
ground motion intensity is likely to
be.

403
00:32:05.170 --> 00:32:08.440
And that results in a hazard curve,
such as using here.

404
00:32:09.540 --> 00:32:14.910
In the mid 1990s when the USGS was
doing this, they were doing these and

405
00:32:14.910 --> 00:32:21.690
developing maps at different return
periods 72 years, 500 years, 1000

406
00:32:21.690 --> 00:32:24.300
years and 2500 years.

407
00:32:29.010 --> 00:32:33.840
When the seismic design values panel
failed to develop a consensus, a

408
00:32:33.840 --> 00:32:38.310
national consensus on how to develop
nationally applicable ground motion

409
00:32:38.310 --> 00:32:44.730
design maps in the US, FEMA, USGS and
BSSC came together again in

410
00:32:44.730 --> 00:32:51.360
1997 with a lever design value panel
that was now called Project 97,

411
00:32:51.960 --> 00:32:57.720
and Project 97 was successful and was
able to come up with what I call the

412
00:32:57.720 --> 00:33:03.030
Goldilocks solution to development of
seismic design maps in the United

413
00:33:03.030 --> 00:33:06.600
States. The Goldilocks solution was not
too warm.

414
00:33:06.630 --> 00:33:10.170
It was not too cold like the porridge
that Goldilocks ate.

415
00:33:10.170 --> 00:33:11.340
It was just right.

416
00:33:13.250 --> 00:33:17.740
The Goldilocks solution consisted of
defining a new design ground motion

417
00:33:17.780 --> 00:33:21.080
level that we call maximum considered
earthquake shaking.

418
00:33:22.520 --> 00:33:28.400
And under the compact that was made in
Project 97 around the United States.

419
00:33:28.640 --> 00:33:33.740
Mce Shaken would be taken as that level
of ground motion that had a 2% 50 year

420
00:33:33.740 --> 00:33:35.210
exceedance probability.

421
00:33:36.530 --> 00:33:42.610
Unless. That motion with 2% 50 year
exceedance probability exceeded

422
00:33:42.610 --> 00:33:48.880
150% of the design ground motion
contained in the 1994 Uniform Building

423
00:33:48.880 --> 00:33:50.800
Code for Zone four.

424
00:33:51.100 --> 00:33:53.720
Now, why 150%?

425
00:33:54.640 --> 00:33:59.320
At one of the meetings of the seismic
design panel, a question was asked.

426
00:34:00.410 --> 00:34:05.780
If you properly design a building today
to the 1994 uniform building code

427
00:34:05.780 --> 00:34:08.780
requirements for Zone four in
California.

428
00:34:09.630 --> 00:34:14.010
How much more intense ground motion do
you think that building could

429
00:34:14.010 --> 00:34:17.370
experience before there would be a
significant risk of the buildings

430
00:34:17.370 --> 00:34:21.110
collapse? And after some thought.

431
00:34:22.080 --> 00:34:25.920
Essentially, everyone in the room
agreed that a well-designed building

432
00:34:26.100 --> 00:34:31.590
for the contemporary uniform building
code could experience probably 150% of

433
00:34:31.590 --> 00:34:34.560
the design ground motion without
collapsing.

434
00:34:36.610 --> 00:34:42.700
And so engineers in the Western US felt
that the seismic design values

435
00:34:42.700 --> 00:34:49.060
contained in the 1994 Uniform Building
Code were adequate for any ground

436
00:34:49.060 --> 00:34:50.620
motion they were likely to see.

437
00:34:51.010 --> 00:34:52.930
And so a compact was reached.

438
00:34:53.470 --> 00:34:58.540
We'll use 2%, 50 year ground motion
everywhere in the United States, except

439
00:34:58.540 --> 00:35:05.140
where that ground motion exceeds 150%
of the 1994 UBC zone four

440
00:35:05.140 --> 00:35:11.550
value. Where are that 2% 50 year motion
exceeded the 1994

441
00:35:11.580 --> 00:35:13.530
UBC zone four value.

442
00:35:14.160 --> 00:35:15.450
We would use.

443
00:35:16.800 --> 00:35:23.310
The lesser of a deterministic motion
equal to 150% of the ground motion

444
00:35:23.310 --> 00:35:27.480
resulting from a characteristic
earthquake on the known active faults

445
00:35:27.480 --> 00:35:32.100
near the site or the 150% UBC value.

446
00:35:32.580 --> 00:35:38.160
And that has basically remained as the
basis for the seismic design maps

447
00:35:38.160 --> 00:35:43.320
contained in the building code since
that time, with a few little tweaks

448
00:35:43.320 --> 00:35:50.270
that I'll talk about. The Project 97
agreement resulted in

449
00:35:50.270 --> 00:35:53.240
maps that look very much like those
we've been used to looking at for the

450
00:35:53.240 --> 00:35:57.530
last 20 years. Separate maps of an S 
sub s parameter and an S sub one

451
00:35:57.530 --> 00:36:02.150
parameter where S sub s and S sub one
represent respectively.

452
00:36:02.750 --> 00:36:07.610
The maximum short period spectral
response acceleration and the maximum

453
00:36:07.610 --> 00:36:13.370
1 second period spectral response
acceleration likely to occur either as

454
00:36:13.370 --> 00:36:18.230
a result of 2% 50 year motion or the
deterministic cap motion.

455
00:36:20.420 --> 00:36:26.450
There are five site classes eight
labeled A, B, C, D, E and F that are

456
00:36:26.450 --> 00:36:31.190
determined based on the expected shear
wave velocity of the soils and the top

457
00:36:31.190 --> 00:36:32.600
30 meters of soil.

458
00:36:33.470 --> 00:36:39.650
And these are these values are tied to
a standard type of spectrum that was

459
00:36:39.650 --> 00:36:44.420
proposed by professors Neumark and Paul
at the University of Illinois in the

460
00:36:44.420 --> 00:36:50.690
1970s. And this standard spectrum has
three domains, a domain of constant

461
00:36:50.690 --> 00:36:52.160
response acceleration.

462
00:36:53.370 --> 00:37:00.010
So that's a zone of constant response
velocity S sub one divided by T

463
00:37:00.060 --> 00:37:05.730
and a zone of constant response to
displacement S sub one divided by T

464
00:37:05.730 --> 00:37:06.360
squared.

465
00:37:13.480 --> 00:37:16.090
The 1997 NEHRP provisions.

466
00:37:16.090 --> 00:37:22.210
Maps were adopted by ASCE 7-98 ASCE
7-02, IBC

467
00:37:22.210 --> 00:37:27.070
2000, IBC 2003, IBC 2006.

468
00:37:28.000 --> 00:37:33.160
But the USGS continued to do
probabilistic seismic risk assessments

469
00:37:33.160 --> 00:37:38.530
and to develop updated contour maps for
adoption by the building codes.

470
00:37:39.190 --> 00:37:43.240
And as they did this, we noticed
significant variation in the ground

471
00:37:43.240 --> 00:37:46.360
motion intensity at different
locations.

472
00:37:47.140 --> 00:37:50.800
Due to increased science that has
occurred over the years, we've learned

473
00:37:50.800 --> 00:37:56.950
about new faults, scientists opinions
as to the magnitude recurrence

474
00:37:56.950 --> 00:38:01.510
relationships have changed over the
years, as have the ground motion

475
00:38:01.510 --> 00:38:05.710
prediction models, and this resulted in
changes in the locations of the

476
00:38:05.710 --> 00:38:07.180
contours of the maps.

477
00:38:07.660 --> 00:38:10.150
And this was troubling to some
engineers.

478
00:38:12.170 --> 00:38:18.860
The maps that were contained in the IBC
2000 census had some difficulties.

479
00:38:18.860 --> 00:38:22.070
They were impossible to read in areas
of high seismicity.

480
00:38:22.730 --> 00:38:28.880
This resulted in the USGS developing a
digital tool that could be used online

481
00:38:28.880 --> 00:38:31.400
to determine the design ground motion
values.

482
00:38:32.470 --> 00:38:36.310
But as I said, as the USGS has
continued to do their probabilistic

483
00:38:36.310 --> 00:38:42.940
hazard studies, the ground motion
values changed and this caused concern.

484
00:38:45.120 --> 00:38:50.730
And so in 2007, the Federal Emergency
Management Agency, the United States

485
00:38:50.730 --> 00:38:57.450
Geological Survey and BSSC funded a
joint project to relook at the way that

486
00:38:57.450 --> 00:39:02.220
the maps were made and see if there
should be any changes.

487
00:39:02.610 --> 00:39:06.570
They did this in light of several
developments.

488
00:39:06.990 --> 00:39:10.800
One, there had been a major project
undertaken by the Pacific Earthquake

489
00:39:10.800 --> 00:39:14.940
Engineering Research Center in which
they had developed new ground motion

490
00:39:14.940 --> 00:39:16.230
prediction models.

491
00:39:17.070 --> 00:39:24.030
These new ground motion prediction
models suggested that in some

492
00:39:24.030 --> 00:39:29.100
regions, ground shaking was going to be
much more intense than had been

493
00:39:29.100 --> 00:39:33.060
previously thought, and in other
regions, much less intense.

494
00:39:33.990 --> 00:39:38.700
The little map that you see over here
on the left shows a representation of

495
00:39:38.730 --> 00:39:45.120
2%, 50 year ground motion intensity or
a spectral response acceleration period

496
00:39:45.120 --> 00:39:46.530
of 1 second.

497
00:39:46.950 --> 00:39:51.990
Comparing the ratio of the values
computed, using the new ground motion

498
00:39:51.990 --> 00:39:57.570
prediction models and the older ones,
and where you see hot colors, yellows

499
00:39:57.570 --> 00:40:01.860
and reds. The ground mode new ground
motion prediction models were

500
00:40:01.860 --> 00:40:04.080
predicting more intense motion.

501
00:40:04.170 --> 00:40:09.690
And where you see colder colors blues,
they were predicting less intense

502
00:40:09.690 --> 00:40:13.740
motion. And in fact, in much of
California, they were predicting

503
00:40:13.740 --> 00:40:19.170
motions that were as low as 70% of
those which had been used for design of

504
00:40:19.170 --> 00:40:21.900
buildings over the past 30 years.

505
00:40:22.690 --> 00:40:27.370
This made engineers in the western
United States uncomfortable.

506
00:40:32.430 --> 00:40:38.640
Engineers on the project 07 worked
closely with USGS and discovered for

507
00:40:38.640 --> 00:40:43.290
the first time that there was
directionality in ground motion,

508
00:40:43.290 --> 00:40:48.300
meaning ground motion isn't as strong
in all three directions of the compass

509
00:40:49.080 --> 00:40:51.420
as it is in other directions.

510
00:40:52.740 --> 00:40:56.640
And also engineers in the eastern
United States were beginning to

511
00:40:56.640 --> 00:40:58.890
complain that we were being forced.

512
00:41:00.160 --> 00:41:07.010
Under the maps developed by BSSC and
USGS and contained in the IBC.

513
00:41:07.010 --> 00:41:11.250
That they were being forced to design
for California like ground motions,

514
00:41:11.520 --> 00:41:14.790
even though they never saw California
like earthquakes.

515
00:41:15.360 --> 00:41:19.560
And so Project 7 tried to deal with
these issues.

516
00:41:21.360 --> 00:41:24.570
Let's talk first about ground motion
directionality.

517
00:41:25.230 --> 00:41:31.680
The plot that you're seeing on the left
is a plot of the acceleration

518
00:41:32.040 --> 00:41:37.620
in the X and Y directions of a real
earthquake as recorded at a particular

519
00:41:37.620 --> 00:41:44.490
site. And what you can see is that
depending on which

520
00:41:44.490 --> 00:41:49.530
direction of the compass you're on, the
ground motion may be substantially more

521
00:41:49.530 --> 00:41:55.850
intense. Or less intense in terms of
the amount of acceleration.

522
00:41:57.140 --> 00:42:02.300
The ground motion prediction models
were using a type of representation of

523
00:42:02.300 --> 00:42:06.950
ground motion intensity known as geo
mean motion.

524
00:42:07.370 --> 00:42:13.790
And what geo mean motion is, is the
square root of the intensity of the

525
00:42:13.790 --> 00:42:19.580
motion in the x recording direction
times that in the Y recording

526
00:42:19.580 --> 00:42:25.070
direction. And for this motion, the geo
mean motion would have a value of the

527
00:42:25.070 --> 00:42:31.850
square root of 0.28 g times 0.5 G or
0.37 g structural

528
00:42:31.850 --> 00:42:34.850
engineers. And looked at this and said,
Well, if you're giving us the geo mean

529
00:42:34.850 --> 00:42:41.030
motion, which is 0.37 G, we know that
the maximum motion at this site was

530
00:42:41.030 --> 00:42:42.590
actually 0.5 G.

531
00:42:42.770 --> 00:42:45.620
So we're not designing for strong
enough motion.

532
00:42:46.650 --> 00:42:50.820
Structural engineers decided that we
need to convert from geometric motion

533
00:42:50.820 --> 00:42:54.030
to so called maximum direction motion.

534
00:42:54.660 --> 00:42:58.860
And based on statistical study would
determine that maximum direction motion

535
00:42:59.760 --> 00:43:04.560
would be about 10% stronger than the
geo mean motion for a short period

536
00:43:04.560 --> 00:43:10.500
spectral response accelerations and
about 30% more for long periods.

537
00:43:10.500 --> 00:43:12.480
Spectral response accelerations.

538
00:43:16.900 --> 00:43:21.670
To respond to the concern by Eastern
engineers that we were forcing them to

539
00:43:21.670 --> 00:43:24.340
design the too large an earthquake.

540
00:43:24.610 --> 00:43:30.100
We looked at the real risk of
experiencing damage in earthquakes over

541
00:43:30.100 --> 00:43:37.050
time. Under ASCE 7-05 and the 2006
International Building

542
00:43:37.050 --> 00:43:41.520
Code Design Values in Memphis,
Tennessee, San Francisco, California

543
00:43:41.520 --> 00:43:43.950
and Los Angeles, California were
similar.

544
00:43:46.320 --> 00:43:49.950
Engineers in the New Madrid region.

545
00:43:49.950 --> 00:43:54.660
Memphis, the St Louis pointed out that
although the design values and the code

546
00:43:54.660 --> 00:43:59.940
were similar in the past 200 years, San
Francisco had experienced five

547
00:43:59.940 --> 00:44:01.290
significant earthquakes.

548
00:44:01.380 --> 00:44:03.630
Los Angeles, eight significant
earthquakes.

549
00:44:03.630 --> 00:44:05.340
And Memphis, only one earthquake.

550
00:44:05.910 --> 00:44:10.650
And it just didn't make sense to them
to design for the same values if

551
00:44:10.650 --> 00:44:12.520
earthquakes happened less often.

552
00:44:12.540 --> 00:44:16.440
It seemed to them that there was less
risk that buildings collapse.

553
00:44:17.530 --> 00:44:19.240
It turns out that they were right.

554
00:44:19.240 --> 00:44:22.960
I don't know that they actually fully
understood why they were right, but

555
00:44:22.960 --> 00:44:28.060
they were right. And so in Project 07,
we looked at this.

556
00:44:28.860 --> 00:44:32.760
We looked at how do you compute the
risk of collapse of a building?

557
00:44:34.440 --> 00:44:37.750
It turns out the way you can compute
the risk of collapse of a building is

558
00:44:37.750 --> 00:44:41.140
with this rather complex equation that
you see here.

559
00:44:41.860 --> 00:44:46.750
You can calculate the number of
collapses you're likely to have a year

560
00:44:46.750 --> 00:44:50.140
in buildings designed to the code by
doing this integral.

561
00:44:50.140 --> 00:44:52.000
And there are two parts that are
integral.

562
00:44:52.570 --> 00:44:56.350
One part is the seismic hazard,
basically the hazard curve that I

563
00:44:56.350 --> 00:45:01.970
showed you earlier. And the other part
is the fragility of the structure.

564
00:45:03.580 --> 00:45:07.810
And about the point in time that this
project was being done.

565
00:45:08.380 --> 00:45:12.670
The Applied Technology Council had done
another project in another FEMA

566
00:45:12.670 --> 00:45:18.880
publication called FEMA P 695 that
presented a methodology for determining

567
00:45:18.880 --> 00:45:20.920
fragility curves for buildings.

568
00:45:22.090 --> 00:45:26.110
And the fragility curve is just a
probability distribution that shows the

569
00:45:26.110 --> 00:45:30.490
probability that a building will
collapse, given that it sees a given

570
00:45:30.490 --> 00:45:32.590
intensity of ground motion.

571
00:45:34.020 --> 00:45:38.310
So if we're a given site with these two
curves, a seismic hazard curve on the

572
00:45:38.310 --> 00:45:42.660
one hand and a fragility curve for a
typical structure, on the other hand,

573
00:45:42.660 --> 00:45:45.870
it's possible to determine the risk of
collapse.

574
00:45:46.320 --> 00:45:51.690
You start by looking at an annual
probability of exceedance for a ground

575
00:45:51.690 --> 00:45:54.840
motion. Figuring out what that ground
motion is.

576
00:45:54.840 --> 00:45:59.040
In this case, I'm taking a 0.01 annual
probability of exceedance.

577
00:45:59.040 --> 00:46:03.450
And for this site I'm finding that the
ground motion intensity has an

578
00:46:03.450 --> 00:46:06.060
acceleration value of 1.5 G.

579
00:46:06.600 --> 00:46:12.840
I then come over into my fragility
curve at 1.5 G and I find that at that

580
00:46:12.840 --> 00:46:17.010
level of ground motion, my fragility
curve tells me there's a 30%

581
00:46:17.010 --> 00:46:18.660
probability of collapse.

582
00:46:19.510 --> 00:46:25.360
For this for this particular ground
motion at this site, I can compute that

583
00:46:25.360 --> 00:46:31.360
there is a 0.001 chance per year of
having the ground motion and a 30%

584
00:46:31.360 --> 00:46:35.350
probability that the building will
collapse if it experiences that ground

585
00:46:35.350 --> 00:46:42.130
motion. And that means that there's a
0.00003 chance per year that a building

586
00:46:42.130 --> 00:46:45.160
will collapse due to 1.5 g ground
motion.

587
00:46:45.760 --> 00:46:50.140
I can repeat this exercise for other
values of ground motion and other

588
00:46:50.140 --> 00:46:56.770
probabilities of exceedance and add
that all up and determine what the

589
00:46:56.770 --> 00:47:02.290
actual total probability of failure of
a building is due to earthquake ground

590
00:47:02.290 --> 00:47:08.370
shaking. Project 07 said that rather
than designing for a uniform risk of

591
00:47:08.370 --> 00:47:14.910
experiencing ground motion of one in
2500 years, instead we would design for

592
00:47:14.910 --> 00:47:18.750
a uniform risk of collapse of 10%.

593
00:47:18.750 --> 00:47:24.240
Given the occurrence of maximum
considered earthquake shaking, which

594
00:47:24.240 --> 00:47:30.690
using approximately 2500 year ground
motion results in a 1% chance in 50

595
00:47:30.690 --> 00:47:33.810
years that a building will collapse in
ground shaking.

596
00:47:34.350 --> 00:47:39.780
So with that and with the adoption of
the Project 07 recommendations, rather

597
00:47:39.780 --> 00:47:45.030
than designing for 2500 year ground
motion, we began designing for ground

598
00:47:45.030 --> 00:47:50.670
motion which produce a 1% 50 year
collapse probability which will result

599
00:47:50.670 --> 00:47:55.440
in sea shaking that has returned
periods varying from maybe 1000 years

600
00:47:55.440 --> 00:48:00.540
in the most seismically active regions
to maybe 3000 years in less active

601
00:48:00.540 --> 00:48:04.440
regions. To implement this into the
code.

602
00:48:04.760 --> 00:48:09.810
USGS developed risk coefficient maps
which had the effect of reducing ground

603
00:48:09.810 --> 00:48:15.030
motion that was designed for in the
eastern US by about 30% and increasing

604
00:48:15.030 --> 00:48:17.280
it slightly in the Western US.

605
00:48:17.730 --> 00:48:22.800
The resulting ground motion maps were
referenced by the 2009 NEHRP provisions

606
00:48:22.950 --> 00:48:29.280
adopted by ASCE 7-10 and 7-16 and by
the 2012, 2015,

607
00:48:29.280 --> 00:48:33.300
2018 and 2021 editions of the
International Building Code.

608
00:48:35.770 --> 00:48:41.980
Late in the process of the 2014 NEHRP
provisions, it was discovered

609
00:48:41.980 --> 00:48:48.730
that the classic new mark and hole
spectral acceleration response

610
00:48:48.730 --> 00:48:54.640
spectrum I showed you earlier doesn't
always work very well, and in

611
00:48:54.640 --> 00:48:58.720
particular it doesn't work well in
sites that are where ground motion is

612
00:48:58.720 --> 00:49:02.650
dominated by large magnitude
earthquakes and the sites have salt

613
00:49:02.650 --> 00:49:08.260
soils. Also, engineers were coming very
unhappy with the pogo stick that I've

614
00:49:08.260 --> 00:49:12.190
described earlier, ground motion values
jumping up and down.

615
00:49:12.520 --> 00:49:19.060
And so FEMA once again established a
new project known as Project 17 because

616
00:49:19.060 --> 00:49:22.510
it occurred in calendar year 2017.

617
00:49:22.510 --> 00:49:28.770
To look at the how the ground motion
maps are generated is a joint BSSC USGS

618
00:49:28.780 --> 00:49:33.850
project. And at the start of the
project we identified a number of

619
00:49:33.850 --> 00:49:38.590
issues in addition to the pogo stick,
the issue of very large magnitude

620
00:49:38.590 --> 00:49:43.210
earthquakes and the different ground
motions that they produce, inclusion of

621
00:49:43.210 --> 00:49:48.940
basin effects and spectral response
ground motion models looking at

622
00:49:48.940 --> 00:49:53.770
spectral shape, the issue of precision
of the ground motion values in the code

623
00:49:53.770 --> 00:49:59.980
versus the uncertainty, acceptable risk
and the use of deterministic CACs.

624
00:50:00.550 --> 00:50:05.920
We went through a process of
eliminating some of those and ended up

625
00:50:05.920 --> 00:50:08.740
with a series of issues that we did
deal with.

626
00:50:10.380 --> 00:50:12.510
First, let's talk about spectral shape.

627
00:50:13.230 --> 00:50:17.460
Graph on the top shows the classic
numeric and whole acceleration response

628
00:50:17.460 --> 00:50:24.150
spectrum. The graph on the right shows
real response spectrum at sites with

629
00:50:24.150 --> 00:50:28.590
similar distance from the cause of
default but having different site

630
00:50:28.590 --> 00:50:34.560
classes. The curves that you see here
in Black, Green and blue represent site

631
00:50:34.560 --> 00:50:41.400
classes A, B, and C and the curves that
you see here in red represent

632
00:50:41.410 --> 00:50:44.720
site class E and an orange site class
D.

633
00:50:45.030 --> 00:50:49.590
And you can see that the spectrum that
we expect from real earthquakes on site

634
00:50:49.590 --> 00:50:54.840
classes A, B and C match pretty well,
the spectral shape from Newmark and

635
00:50:54.840 --> 00:50:58.890
Hall. But on site classes D and E, they
do not.

636
00:50:59.790 --> 00:51:02.070
And so it was decided to deal with
that.

637
00:51:02.430 --> 00:51:05.100
We couldn't find a good way to deal
with that.

638
00:51:05.130 --> 00:51:11.790
We put a patch into the 2016 edition of
ASCE 7 that said You have to do a site

639
00:51:11.790 --> 00:51:12.990
specific study.

640
00:51:13.080 --> 00:51:16.110
If you're on one of these sites that
can experience large magnitude

641
00:51:16.110 --> 00:51:20.250
earthquakes and you have soft soils,
but that wasn't particularly

642
00:51:20.250 --> 00:51:27.030
satisfying. So under Project 2017, it
was decided that USGS, instead of

643
00:51:27.030 --> 00:51:33.540
publishing maps of subsets as of one
values, instead would publish

644
00:51:33.540 --> 00:51:38.340
maps of the spectral response
acceleration at 20 different periods,

645
00:51:38.340 --> 00:51:43.080
ranging from 0 seconds up to 10
seconds, that these values would be

646
00:51:43.080 --> 00:51:45.480
automatically site class adjusted.

647
00:51:45.810 --> 00:51:50.190
And because there were big differences
between site class C and D and site

648
00:51:50.190 --> 00:51:56.590
class D and E that we would establish
some new site classes BC, CD, DE

649
00:51:57.060 --> 00:52:03.270
being at these best values intermediate
to those for C, d and E.

650
00:52:04.290 --> 00:52:09.420
Because USGS was now computing site
adjusted values and spectral response

651
00:52:09.420 --> 00:52:14.160
acceleration. The values of EPS of A
and that's a V formerly in the building

652
00:52:14.160 --> 00:52:19.200
code disappeared as did the values of
S sub s and S sub one.

653
00:52:20.860 --> 00:52:23.950
Now, this created somewhat of a dilemma
for us.

654
00:52:25.030 --> 00:52:28.390
In 1991, the building code had one map.

655
00:52:28.880 --> 00:52:32.050
In the 2000 IBC, it had 14 maps.

656
00:52:32.380 --> 00:52:37.200
In the 2000 IBC, 20 maps, 2010, 32
maps.

657
00:52:37.210 --> 00:52:42.340
And if we had published maps for each
of these different site classes, we

658
00:52:42.340 --> 00:52:47.680
would have ended up with something like
2000 maps in the in the building code,

659
00:52:47.680 --> 00:52:49.480
which was clearly unworkable.

660
00:52:50.080 --> 00:52:54.760
And so we've eliminated maps from the
present edition of the provisions and

661
00:52:54.760 --> 00:53:01.270
ASCE seven and instead direct you to an
online tool maintained by ASCEA

662
00:53:01.420 --> 00:53:05.500
that grabs data from the USGS database.

663
00:53:06.790 --> 00:53:11.620
And so with that, we're ready to go
back to the future and I'll turn the

664
00:53:11.620 --> 00:53:14.080
floor over to Sanaz.

665
00:53:17.510 --> 00:53:23.850
Thank you, Ron. That was a very nice
overview of a very long history and I'm

666
00:53:23.850 --> 00:53:30.330
going to try to share my screen right
now, so I apologize for a minute of

667
00:53:30.330 --> 00:53:36.970
transition. Okay.

668
00:53:37.320 --> 00:53:43.320
Hi, everyone. I'm Sanaz Rezaeian, and
I'm a researcher, structural engineer

669
00:53:43.530 --> 00:53:45.690
with the US Geological Survey.

670
00:53:46.830 --> 00:53:49.320
I hope that you can see my screen right
now.

671
00:53:50.610 --> 00:53:54.840
If you cannot, please let me know,
Jiqiu or Ron.

672
00:53:55.710 --> 00:53:56.460
Yes, I can.

673
00:53:57.150 --> 00:53:58.260
Great. Thank you.

674
00:53:58.290 --> 00:54:04.830
So I'm going to cover a part of chapter
three in this training

675
00:54:04.830 --> 00:54:09.330
material that NIBS has put together for
the users of 2020 NEHRP.

676
00:54:10.230 --> 00:54:15.510
And this part of chapter three
basically goes over the updated USGS

677
00:54:15.690 --> 00:54:22.140
hazard model. To be specific, this is a
2018 USGS hazard model that uses the

678
00:54:22.140 --> 00:54:24.690
best available science and data to
date.

679
00:54:24.690 --> 00:54:28.290
And that's what Ron was setting up the
stage for.

680
00:54:28.650 --> 00:54:35.220
So we basically moved into 2018, and
it's from the point that he stopped

681
00:54:35.820 --> 00:54:40.140
presenting. So I will go over the
details of this a little bit.

682
00:54:40.710 --> 00:54:44.280
First, here is the outline of my
presentation.

683
00:54:45.180 --> 00:54:51.060
I'm going to very quickly go over the
interplay between the USGS hazard

684
00:54:51.060 --> 00:54:54.960
models and the BSSC PUC requirements.

685
00:54:55.260 --> 00:55:00.360
Basically, at the USGS we calculate
design ground motions, but they are a

686
00:55:00.360 --> 00:55:04.830
combination of the probabilistic ground
motions that are calculated by the US

687
00:55:04.830 --> 00:55:11.580
Geological Survey and also combined
with the BSSC PUC procedures.

688
00:55:11.580 --> 00:55:15.810
And these are things that Ron went over
and can include things like risk,

689
00:55:15.820 --> 00:55:20.400
targeted motion calculations,
application of site amplifications,

690
00:55:20.400 --> 00:55:24.720
deterministic capping of those
probabilistic ground motions, addition

691
00:55:24.720 --> 00:55:29.310
of maximum direction factors to
basically arrive at what is commonly

692
00:55:29.310 --> 00:55:33.690
referred to as the risk targeted,
maximum considered earthquake shaking.

693
00:55:34.050 --> 00:55:38.790
So we do calculate these design ground
motions at the USGS, but they are a

694
00:55:38.790 --> 00:55:45.340
combination of science that USGS builds
and policy that are said by the CCP.

695
00:55:46.190 --> 00:55:51.450
You see, then I would like to spend
most of my time with you going into the

696
00:55:51.450 --> 00:55:57.210
details of the 2018 USGS national
seismic hazard model that we refer to

697
00:55:57.210 --> 00:55:58.830
as Inception, for sure.

698
00:55:59.400 --> 00:56:03.150
What's important to note here is that
this update was only for the

699
00:56:03.150 --> 00:56:04.560
contaminant for us.

700
00:56:04.590 --> 00:56:10.500
This is the lower 48 states and it
included two major

701
00:56:10.740 --> 00:56:16.320
improvements. One was that we
implemented all new ground motion

702
00:56:16.320 --> 00:56:19.410
models in the central and eastern U.S.

703
00:56:19.410 --> 00:56:23.280
or see us, including in the east
models.

704
00:56:24.390 --> 00:56:29.310
The other major improvement was that we
incorporated deep basin effects in the

705
00:56:29.310 --> 00:56:30.570
western U.S.

706
00:56:31.110 --> 00:56:32.310
or w us.

707
00:56:32.310 --> 00:56:36.090
And I will go over the detail of these
two improvements.

708
00:56:36.480 --> 00:56:41.850
I will also spend just a little bit of
time like 5 minutes at the end letting

709
00:56:41.850 --> 00:56:44.730
you know what we did outside of
containment is us.

710
00:56:44.760 --> 00:56:51.210
These are locations in Hawaii, Alaska,
Puerto Rico, Guam and Northern Mariana

711
00:56:51.210 --> 00:56:52.890
Islands and American Samoa.

712
00:56:52.890 --> 00:56:57.150
Was that basically USGS hasn't updated
the hazard models to provide

713
00:56:57.150 --> 00:56:59.240
multivariate responsive spectrum yet.

714
00:56:59.490 --> 00:57:04.230
So I'm going to let you know what we
did to arrive at all the ground motions

715
00:57:04.230 --> 00:57:07.020
that were needed for O'Connor's
locations.

716
00:57:08.190 --> 00:57:09.600
So let's get started.

717
00:57:10.140 --> 00:57:14.550
In this table, you can see that USGS
national seismic hazard models have

718
00:57:14.550 --> 00:57:19.680
been used in calculation of design
ground motions in major provisions.

719
00:57:19.890 --> 00:57:24.810
Then a seven standards and then
international building codes.

720
00:57:24.810 --> 00:57:30.720
Since 1996, USGS updated these systems
in

721
00:57:30.750 --> 00:57:37.260
2002, 2008, 2014, and the latest update
in the US has been in

722
00:57:37.260 --> 00:57:41.340
2018, as Ron mentioned before.

723
00:57:41.580 --> 00:57:47.340
Basically what we do, our scientists do
at the USGS is to calculate

724
00:57:48.630 --> 00:57:52.050
hazard curves at every grid point on
the map.

725
00:57:52.200 --> 00:57:57.360
And this is done by performing
probabilistic seismic hazard analysis

726
00:57:57.360 --> 00:58:00.870
or passage at every point on the US
map.

727
00:58:02.590 --> 00:58:08.200
Our Engineering and risk project group
takes these hazard curves and combines

728
00:58:08.200 --> 00:58:13.460
them with site specific procedures that
are specified in Chapter 21 of near

729
00:58:13.480 --> 00:58:18.910
provisions. And this is where the CPC
has specified the site specific

730
00:58:18.910 --> 00:58:23.320
procedures to include things like risk
targeted ground motion calculations,

731
00:58:23.320 --> 00:58:28.440
addition of maximum direction factors,
site amplifications and deterministic

732
00:58:28.450 --> 00:58:33.880
mapping. And what we do is to calculate
design ground motion maps in the past

733
00:58:33.880 --> 00:58:36.880
and now just design ground motion data
database.

734
00:58:38.450 --> 00:58:44.630
In the previous cycles, USGS used to
provide hazard curves and gram

735
00:58:44.630 --> 00:58:51.560
motions at three periods of PGA point
to second and 1/2 and at one

736
00:58:51.560 --> 00:58:57.050
reference site class that was specified
with VS 3760 meters per second.

737
00:58:57.590 --> 00:59:03.560
In the latest version, we are actually
providing ground motions for 22 periods

738
00:59:03.950 --> 00:59:09.350
all the way from PGA to 10 seconds and
for eight different site classes.

739
00:59:09.440 --> 00:59:13.670
And I say eight versus the nine that
are mentioned because we're not

740
00:59:14.030 --> 00:59:15.650
providing this for site class.

741
00:59:16.340 --> 00:59:19.020
So this is five classes A through E.

742
00:59:20.620 --> 00:59:25.390
And as you can see, the latest
promotions based on the 2018 US jazz

743
00:59:25.450 --> 00:59:31.870
hazard model have been adopted in the
2020 NE Herb and also in AC 722 right

744
00:59:31.870 --> 00:59:38.190
now. And they are being proposed and
considered for 2024 IBC code.

745
00:59:39.720 --> 00:59:43.980
So it's going to the changes that have
happened to the design promotions,

746
00:59:44.010 --> 00:59:47.790
basically all the updates that you will
see in 2020.

747
00:59:47.880 --> 00:59:51.780
We have design promotions in the
content business as they come from two

748
00:59:51.780 --> 00:59:54.960
different sources of improvements and
changes.

749
00:59:55.350 --> 01:00:00.510
One source of changes come from the
2018 USGS National Seismic Hazard

750
01:00:00.510 --> 01:00:03.290
model. That updated earthquake sources.

751
01:00:03.300 --> 01:00:08.340
These are earthquake magnitudes, rates
of earthquakes, happening faults zones,

752
01:00:08.340 --> 01:00:12.540
things like that, and also updated
promotion models.

753
01:00:12.570 --> 01:00:15.960
These are the old gaps that we call
GMES now.

754
01:00:16.980 --> 01:00:23.490
Another source of changes is the CC
Project 17 recommendations to update

755
01:00:23.490 --> 01:00:26.520
the site to specific procedures of
Chapter 21.

756
01:00:27.420 --> 01:00:31.650
So I'm going to very briefly summarize
these recommendations for you, but

757
01:00:31.650 --> 01:00:32.760
there is a lot more to it.

758
01:00:32.760 --> 01:00:36.960
And I encourage you to actually read
these training materials and look at

759
01:00:36.960 --> 01:00:41.430
2029 provisions in the commentary
chapters for more detail.

760
01:00:42.450 --> 01:00:47.850
In short, what Project 17 recommended
was not to change the targeted

761
01:00:47.870 --> 01:00:49.370
promotion calculations.

762
01:00:49.380 --> 01:00:52.710
Obviously, this is not a change, but
it's really important because a group

763
01:00:52.710 --> 01:00:56.710
of scientists and engineers actually
looked into this issue.

764
01:00:56.730 --> 01:01:00.780
They considered whether we should go
back to these seismic zone warnings or

765
01:01:00.780 --> 01:01:04.890
change the risk, the target risk or a
lot of other things.

766
01:01:04.890 --> 01:01:08.720
And they decided that we should
actually stick with the current risk

767
01:01:08.730 --> 01:01:10.770
target calculations that we have.

768
01:01:11.820 --> 01:01:16.140
They did recommend to use multi period
and multi-year 30 responses background

769
01:01:16.140 --> 01:01:21.660
that for short we call multi period
response a spectrum where spectral is

770
01:01:21.680 --> 01:01:23.820
plural for several years thirties.

771
01:01:25.290 --> 01:01:30.270
They also recommended that we should
modify deterministic capping

772
01:01:30.270 --> 01:01:36.120
calculations, which is now based on
this aggregation of probabilistic

773
01:01:36.120 --> 01:01:40.770
hazard instead of the characteristic
earthquakes that they used to have.

774
01:01:41.570 --> 01:01:45.470
And finally they recommended to update
the maximum direction factors to be

775
01:01:45.470 --> 01:01:47.570
based on new data and studies.

776
01:01:48.350 --> 01:01:52.790
Now I'm going to focus in the rest of
this presentation on this issue of

777
01:01:52.790 --> 01:01:57.140
multi period response as background,
because this actually influenced our

778
01:01:57.770 --> 01:02:02.600
update of the USGS national seismic
hazard model in 2018.

779
01:02:02.720 --> 01:02:06.500
And that was because we had to make
sure the ground motion models that we

780
01:02:06.500 --> 01:02:11.690
use where applicable for all the
periods and classes that were required

781
01:02:11.690 --> 01:02:14.000
by multivariate response spectrum.

782
01:02:15.320 --> 01:02:21.530
So the major changes in 2018 USGS
hazard model included using new ground

783
01:02:21.530 --> 01:02:24.620
motion models in the central and
eastern US.

784
01:02:24.740 --> 01:02:31.370
This includes new motion models and a
new side amplification factor model

785
01:02:31.370 --> 01:02:33.740
that was specific to the US.

786
01:02:35.060 --> 01:02:39.740
We incorporated deep basin effects in
four regions in the western U.S.

787
01:02:39.770 --> 01:02:44.570
being Los Angeles, Seattle, San
Francisco and Salt Lake City Basins.

788
01:02:45.950 --> 01:02:51.110
We also made some minor, relatively
minor modifications to commercial

789
01:02:51.110 --> 01:02:54.710
models for crustal and subduction zone
earthquakes.

790
01:02:55.160 --> 01:02:59.810
And finally, we updated our background
seismicity to include more recent

791
01:02:59.810 --> 01:03:02.390
recordings of earthquakes in the US.

792
01:03:03.650 --> 01:03:07.820
What's important to note here is that
the first three updates were actually

793
01:03:07.820 --> 01:03:13.430
necessary for us to be able to provide
multiple period response, a spectrum

794
01:03:13.430 --> 01:03:17.420
that was being requested by the b c
project 17.

795
01:03:18.390 --> 01:03:21.570
So let's get into the details of these
changes.

796
01:03:21.570 --> 01:03:25.380
As I mentioned, the updated all the
Graham motion models in the central and

797
01:03:25.380 --> 01:03:30.720
eastern U.S. and what you see here are
the older brand motion models that we

798
01:03:30.720 --> 01:03:33.330
had in 2014 version of.

799
01:03:34.740 --> 01:03:40.500
Basically, what we had before was very
limited in terms of periods and side

800
01:03:40.500 --> 01:03:44.310
classes, and they all had to be
replaced.

801
01:03:45.820 --> 01:03:52.000
Fortunately, around the same time in
Jamie's project that was coordinated by

802
01:03:52.000 --> 01:03:57.640
Pacific Earthquake Engineering Research
Center was completed, and it provided

803
01:03:57.640 --> 01:04:03.160
us with the resources we needed to
update motion models in the US.

804
01:04:03.760 --> 01:04:07.180
We ended up using two sets of models.

805
01:04:07.330 --> 01:04:13.330
One was a set of 14 updated seat
promotion models that were developed by

806
01:04:13.330 --> 01:04:19.180
individual modelers, and we gave that
group a rate of one third.

807
01:04:19.810 --> 01:04:26.740
We also use the final 17 MGM Grand
motion models that

808
01:04:26.740 --> 01:04:32.440
came out of a Sammons mapping
procedure, which took as input these

809
01:04:32.710 --> 01:04:34.030
grand motion models.

810
01:04:34.030 --> 01:04:38.710
And the goal of this set of models was
to better represent epistemic

811
01:04:38.710 --> 01:04:41.140
uncertainties in the grand motion
space.

812
01:04:41.290 --> 01:04:47.050
And at the USGS, we gave this group a
rate of two thirds, basically ending up

813
01:04:47.050 --> 01:04:51.220
with 31 new ground motion models
compared to what we had in the previous

814
01:04:51.220 --> 01:04:57.090
cycle. What all of these motion models
are now applicable to the entire period

815
01:04:57.100 --> 01:05:00.490
range from 8 to 10 second that we
needed.

816
01:05:01.830 --> 01:05:07.050
In general, we saw some changes to the
media and ground motion values in

817
01:05:07.050 --> 01:05:09.990
general increases for large magnitude,
earthquakes.

818
01:05:09.990 --> 01:05:12.020
That's New Madrid seismic zone.

819
01:05:12.030 --> 01:05:14.360
From middle to large distances.

820
01:05:14.370 --> 01:05:18.990
We saw increased epistemic uncertainty,
which was something to be expected in

821
01:05:18.990 --> 01:05:21.150
this particular cycle.

822
01:05:21.150 --> 01:05:24.630
We saw little changes to obligatory
uncertainty.

823
01:05:24.630 --> 01:05:29.490
And here you see the actual or
emotional models that we ended up

824
01:05:29.490 --> 01:05:31.440
using. We have the two sets.

825
01:05:31.440 --> 01:05:36.870
There are advantages and disadvantages
to both sets, and that's how we

826
01:05:36.870 --> 01:05:39.720
determine that one third and two third
rate.

827
01:05:41.410 --> 01:05:47.230
One thing to note about all these new
promotional models is that they are all

828
01:05:47.260 --> 01:05:52.840
developed for very hard rock site
conditions with a study of 3000 meters

829
01:05:52.840 --> 01:05:59.200
per second. There was a working group
that started with Angus and that was

830
01:05:59.200 --> 01:06:04.660
led by John Stewart and others, and
this group came up with a side effect

831
01:06:04.660 --> 01:06:07.120
model for central and eastern US.

832
01:06:08.650 --> 01:06:11.170
The side effect model has three terms
in it.

833
01:06:11.170 --> 01:06:18.040
There is the F 760 term that transforms
the 3000 meter per second spectrum to

834
01:06:18.040 --> 01:06:20.820
760 reference conditions.

835
01:06:20.830 --> 01:06:26.860
Then there is a linear amplification
and a non-linear amplification to each

836
01:06:26.860 --> 01:06:33.030
of which are described in much detail
in two papers in earthquake spectrum.

837
01:06:33.580 --> 01:06:38.170
Usgs interacted with these groups quite
a bit and that's why you see an example

838
01:06:38.170 --> 01:06:44.620
of one of our memos with this group,
and we basically ended up implementing

839
01:06:44.620 --> 01:06:49.420
their suggested model with slight
changes, and that's the differences.

840
01:06:49.420 --> 01:06:55.360
You see these dotted and solid lines
right around 0.1 second and 1/2.

841
01:06:55.870 --> 01:07:00.310
All of these changes are documented in
a paper that we have in earthquake

842
01:07:00.400 --> 01:07:05.710
spectrum. What's important to note here
is that this is the first time that UCG

843
01:07:05.710 --> 01:07:10.930
is actually used as a side effect model
that's specific to the central and

844
01:07:10.930 --> 01:07:12.520
eastern US.

845
01:07:12.610 --> 01:07:19.270
Previously, the if A and B factors were
actually based on Western US data, but

846
01:07:19.270 --> 01:07:24.190
they were being applied to the entire
US and we knew that was not the ideal,

847
01:07:24.430 --> 01:07:25.690
ideal case.

848
01:07:25.690 --> 01:07:30.760
So this in itself is a major
improvement for the region.

849
01:07:32.040 --> 01:07:37.740
Now in this slide, what I'm showing to
you is the ratio between our 2018 model

850
01:07:37.740 --> 01:07:43.140
to the 2014 model, basically showing
that our new commercial motion models

851
01:07:43.140 --> 01:07:48.120
in the East caused an increased ring
around the New Madrid seismic zone, as

852
01:07:48.120 --> 01:07:49.560
I was mentioning before.

853
01:07:49.860 --> 01:07:55.500
And this is because our medians went up
in this region, but also epistemic

854
01:07:55.500 --> 01:08:01.680
uncertainty increase a little more
right around this ring for the really

855
01:08:01.710 --> 01:08:07.440
for the smaller magnitude events and in
larger distances from large magnitude

856
01:08:07.440 --> 01:08:14.010
events, we do see some decreases right
around New York area, East Coast kind

857
01:08:14.010 --> 01:08:14.070
of.

858
01:08:14.070 --> 01:08:14.520
Thing.

859
01:08:15.510 --> 01:08:19.860
In this map. One more thing that you
can note is that there are some

860
01:08:19.860 --> 01:08:24.930
decreases and some localized increases
in the intermountain west region, and

861
01:08:24.930 --> 01:08:31.890
that's because of this new seismicity
catalog that we use in the 2018 update.

862
01:08:34.010 --> 01:08:39.770
Second major improvement in the 2018 US
gas hazard model was that we

863
01:08:39.770 --> 01:08:44.150
incorporated deep bass and effects in
the four regions that you see on the

864
01:08:44.150 --> 01:08:50.540
map in Seattle, Bay Area, Los Angeles
and Wasatch Basin area.

865
01:08:51.830 --> 01:08:57.620
And that's because we had reliable
measurements of parameter Z one and Z

866
01:08:57.620 --> 01:08:59.600
2.5 in these regions.

867
01:08:59.990 --> 01:09:05.060
These are parameters that we use to
categorize the base in depth terms.

868
01:09:05.210 --> 01:09:12.110
And the way that USGS implemented these
in 2018 was basically to use

869
01:09:12.110 --> 01:09:16.370
measurements in the deeper portions of
the basin that you see in the darker

870
01:09:16.370 --> 01:09:23.000
color, but use default values in the
shallower parts of the basin and also

871
01:09:23.000 --> 01:09:24.250
outside of the basin.

872
01:09:24.260 --> 01:09:29.570
So kind of having a smooth transition
going over these basin boundaries that

873
01:09:29.570 --> 01:09:32.090
are kind of arbitrary drawn.

874
01:09:34.340 --> 01:09:39.080
What you see in this slide is the
changes we had to make to our grand

875
01:09:39.080 --> 01:09:45.440
motion models in the western U.S., to
crustal motion models and subduction

876
01:09:45.440 --> 01:09:50.750
grab motion models to basically
incorporate these basin deep basin

877
01:09:50.750 --> 01:09:57.200
amplifications. The implementation of
USGS is slightly different from the

878
01:09:57.200 --> 01:10:01.460
definition of grand motion models as
they have been published, and that's

879
01:10:01.460 --> 01:10:07.790
because we wanted to make sure that the
amplifications are fully applied at 1/2

880
01:10:07.790 --> 01:10:10.700
and periods that are longer.

881
01:10:11.120 --> 01:10:15.430
But we didn't want the Basin
Amplifications to change the motions at

882
01:10:15.470 --> 01:10:21.350
0.5 second periods and shorter periods
with subduction graham motion models.

883
01:10:21.350 --> 01:10:26.000
We had another set of problems and that
was because these grand motion models

884
01:10:26.000 --> 01:10:27.140
that we had B.C.

885
01:10:27.170 --> 01:10:32.710
Hydro, Atkinson, Macias and Shell, they
didn't have a baseline determining

886
01:10:33.620 --> 01:10:37.430
what. It was pretty obvious to us that
something should be included for the

887
01:10:37.430 --> 01:10:38.660
Seattle basin.

888
01:10:39.200 --> 01:10:44.570
So what we ended up doing was to modify
these grand motion models for longer

889
01:10:44.570 --> 01:10:50.510
periods by using Campbell and
Bozorgmehr 2014 model, just to make

890
01:10:50.510 --> 01:10:55.940
sure that we're not underestimating the
motion or that we're providing hazard

891
01:10:55.940 --> 01:10:58.520
at long periods beyond 1/2.

892
01:10:59.820 --> 01:11:05.280
What you see in this slide is the
effect of this deep basin effect being

893
01:11:05.520 --> 01:11:07.260
applied in these regions.

894
01:11:07.680 --> 01:11:12.840
These maps right here are showing
ground motions at a long period of 5/2

895
01:11:12.840 --> 01:11:14.760
and a soft slide class DX.

896
01:11:14.850 --> 01:11:20.400
And what they show is the actual values
that we ended up having in 2018 using

897
01:11:20.400 --> 01:11:27.390
local basin depth divided by what we
would have had in 2018 if we

898
01:11:27.390 --> 01:11:30.480
use default basin depth.

899
01:11:30.720 --> 01:11:35.160
Basically, if you look at Seattle right
here, what this is telling you is that

900
01:11:35.160 --> 01:11:39.540
the ground motion is twice as large as
it would have been if we had used

901
01:11:39.560 --> 01:11:44.940
default values based on these 30 values
for the site of interest.

902
01:11:45.360 --> 01:11:49.680
And these are more or less consistent
with recent studies and simulations

903
01:11:49.680 --> 01:11:54.600
that have been done in the Seattle
region and some in the Los Angeles

904
01:11:54.600 --> 01:12:01.410
region. Now, everything that I talked
about was about our

905
01:12:01.410 --> 01:12:07.890
2018 inches system update, which is
just for the terminus us for locations

906
01:12:07.890 --> 01:12:09.940
outside of Terminus us.

907
01:12:09.960 --> 01:12:13.350
As I mentioned before, like Hawaii,
Alaska, Puerto Rico.

908
01:12:14.040 --> 01:12:19.890
Our hazard models were not updated, so
our older hazard models only provided

909
01:12:19.890 --> 01:12:26.430
values of ss1 and pt l map that run
also mentioned in his presentation

910
01:12:26.430 --> 01:12:33.300
briefly. So we developed a method for
these regions based on what we saw in

911
01:12:33.300 --> 01:12:40.290
Western us, in Terminus, us to
basically come up with

912
01:12:40.290 --> 01:12:45.630
these generic spectral shapes that are
normalized at s value at point to

913
01:12:45.630 --> 01:12:48.500
second and b c cycles.

914
01:12:48.600 --> 01:12:55.590
And then there are guidelines in this
female P 2078 report

915
01:12:55.590 --> 01:13:02.220
that specify how the user can scale
these generic shapes for short

916
01:13:02.220 --> 01:13:08.670
periods to match the assessed value of
the site and scale them at long periods

917
01:13:08.670 --> 01:13:14.640
to match as one value of the site and
then blend the two to determine what

918
01:13:15.210 --> 01:13:19.110
the grab motion should be for mid
period ranges.

919
01:13:20.490 --> 01:13:25.590
All of these are outlined in much
detail in this report and the report is

920
01:13:25.590 --> 01:13:31.230
actually free for download on the NBS
website, so I invite you to look at it.

921
01:13:32.590 --> 01:13:37.600
We validated this approach for
locations where we had actual values of

922
01:13:37.600 --> 01:13:41.950
multivariate response spectrums, such
as in Irvine, California, and the left

923
01:13:41.950 --> 01:13:46.150
hand side FIG.. What you see in solid
lines are our predictions.

924
01:13:46.150 --> 01:13:52.030
Just by matching assigned values and in
dotted lines or dashed lines, you can

925
01:13:52.030 --> 01:13:58.870
see the actual dnpr is being calculated
using 2018 US GIs in a section

926
01:13:59.740 --> 01:14:03.940
after we were satisfied with the
performance of this procedure, which as

927
01:14:03.940 --> 01:14:07.480
you can see, is actually really good at
this location.

928
01:14:07.600 --> 01:14:13.920
We took that method and predicted the
entire spectrum of multi period

929
01:14:14.080 --> 01:14:20.320
response spectrum for locations outside
of Terminus us, such as this one in

930
01:14:20.320 --> 01:14:21.700
Honolulu, Hawaii.

931
01:14:21.700 --> 01:14:26.410
So what you see in the right hand side
figure matches the is and is one values

932
01:14:26.410 --> 01:14:29.740
of the 1996 hazard model in Hawaii.

933
01:14:30.310 --> 01:14:34.510
What everything else is predicted based
on Western US data.

934
01:14:35.930 --> 01:14:41.660
In summary, the multi period response a
spectral requirement of the CPC

935
01:14:41.660 --> 01:14:47.900
influence. Our 2018 update of the USGS
National Seismic Hazard model

936
01:14:48.290 --> 01:14:53.240
mainly because we had to make sure our
promotion models were applicable for

937
01:14:53.240 --> 01:14:58.180
all the periods and classes of interest
and where they were not applicable.

938
01:14:58.190 --> 01:15:00.200
We had to make changes to them.

939
01:15:02.360 --> 01:15:08.750
The updates that included in the 2018
USGS maps models

940
01:15:08.750 --> 01:15:14.240
included new ocean models in the
central and eastern U.S., where we had

941
01:15:14.240 --> 01:15:19.880
14 updated seed models, 17 inches
models and a new side effect model.

942
01:15:20.300 --> 01:15:24.830
They also included incorporation of
deep basin effects in four regions in

943
01:15:24.830 --> 01:15:26.090
the western U.S..

944
01:15:26.330 --> 01:15:33.050
Removal of one crustal GMM and change
in modification of

945
01:15:33.050 --> 01:15:37.340
subduction zones which are relatively
minor and even over very quickly.

946
01:15:37.880 --> 01:15:43.160
And finally, we updated the seismicity
catalog, mostly making changes to the

947
01:15:43.160 --> 01:15:45.350
intermountain west region.

948
01:15:46.650 --> 01:15:52.530
All of these are detailed in a series
of five earthquake spectrum papers that

949
01:15:52.530 --> 01:15:54.120
I want you to look at.

950
01:15:54.120 --> 01:15:58.950
If you're more curious and you're more
than welcome to email myself or any of

951
01:15:58.950 --> 01:16:04.620
these authors, and we will help you to
find these papers, which are, by the

952
01:16:04.620 --> 01:16:08.580
way, free for download because we made
them open source.

953
01:16:08.580 --> 01:16:13.590
Since this time we didn't have an open
file report for our update.

954
01:16:14.010 --> 01:16:18.050
And finally, four locations outside of
Terminus us.

955
01:16:18.060 --> 01:16:23.880
We followed the procedure of female P
2078 report basically approximating

956
01:16:23.880 --> 01:16:30.030
multivariate response, a spectrum only
known as ss1 and rt l values.

957
01:16:30.180 --> 01:16:32.010
For all these locations.

958
01:16:32.010 --> 01:16:37.530
We do expect USGS to update the hazard
models in the coming years, which then

959
01:16:37.530 --> 01:16:39.150
will replace these values.

960
01:16:39.150 --> 01:16:41.100
But for now, what's in 2020?

961
01:16:41.100 --> 01:16:44.340
Ne provision is based on these
estimation.

962
01:16:46.210 --> 01:16:47.670
So that's all I had.

963
01:16:47.680 --> 01:16:51.010
And I want to thank you all for
sticking with us and listening to these

964
01:16:51.010 --> 01:16:55.480
presentations. And now I think we're
going to take questions not just for

965
01:16:55.480 --> 01:17:01.240
me, but also for Ron, since we
transitioned and you didn't have a

966
01:17:01.240 --> 01:17:02.740
chance to ask him a question.

967
01:17:02.950 --> 01:17:07.130
Gq Thank you, thank you, thank you,
thank you.

968
01:17:07.420 --> 01:17:10.320
Thank you again for the great
presentation.

969
01:17:10.330 --> 01:17:13.570
I think in the chat, in the Q&A, we got
a lot of applause.

970
01:17:13.840 --> 01:17:15.850
I really appreciate you to be here
today.

971
01:17:15.970 --> 01:17:18.230
So we do get a lot of questions.

972
01:17:18.250 --> 01:17:23.080
I think everyone has answered most,
almost all of them in the Q&A box.

973
01:17:23.650 --> 01:17:24.850
Please check there.

974
01:17:24.880 --> 01:17:26.860
There is an open question column.

975
01:17:26.860 --> 01:17:28.600
There is always answered question.

976
01:17:28.780 --> 01:17:32.410
So make sure to check there for some of
those questions that Iran has already

977
01:17:32.410 --> 01:17:38.380
answered. Also, I think we have a few
minutes for more questions before we do

978
01:17:38.380 --> 01:17:42.520
that. Do you want to share the link for
this webinar series?

979
01:17:43.120 --> 01:17:47.470
So as I mentioned the very beginning,
this is the 11 webinar series for the

980
01:17:47.470 --> 01:17:49.210
new design examples.

981
01:17:50.170 --> 01:17:52.780
This is our first one on the motions.

982
01:17:52.780 --> 01:17:55.520
So our next one is on June 2nd.

983
01:17:55.990 --> 01:18:00.640
You can check the link at us here in
the chat will be given by Charlie

984
01:18:00.640 --> 01:18:06.640
Kershaw, Seb Krauss and Nico Loco to
talk about some of those changes that

985
01:18:06.640 --> 01:18:13.570
was shared today based on the multiple
response to all the changes introduced

986
01:18:13.570 --> 01:18:17.830
in the last cycle. So make sure check
that you can register for the next one.

987
01:18:17.860 --> 01:18:21.820
So with that, let's see if we have more
questions.

988
01:18:35.760 --> 01:18:38.640
I see now. I mean, if there are no more
questions, Ron, thank you.

989
01:18:38.660 --> 01:18:40.690
You are very efficient to me.

990
01:18:40.740 --> 01:18:46.540
All those answers again, feel free to
reach out to myself to run or not.

991
01:18:47.030 --> 01:18:50.430
If if you have more questions or I'll
have one more coming.

992
01:18:52.330 --> 01:18:55.360
Thank you again.

993
01:18:55.630 --> 01:18:58.910
Let's give our listeners another
virtual round of applause.

994
01:18:58.930 --> 01:19:00.970
Thank you again to being here with us
also.

995
01:19:00.970 --> 01:19:05.170
Thank you, all of you, to be here with
us today to listen to these webinars.

996
01:19:05.200 --> 01:19:10.270
With that, I will close the webinar for
the day and we'll see you guys next

997
01:19:10.270 --> 01:19:12.620
time. Thank you.

998
01:19:13.120 --> 01:19:14.540
Very much. Thank you.