WEBVTT

00:00:00.000 --> 00:00:03.000
Oh, okay.

00:00:03.000 --> 00:00:10.000
Okay, welcome back, everyone, to the last, but definitely not least.

00:00:10.000 --> 00:00:25.000
session of day one of our walk shop. Um, we are now going to talk about before, during, and after the earthquake, the latest in retrofits, loss modeling, and recovery, and to that end, we want to welcome our moderators, Cheryl Movinovich.

00:00:25.000 --> 00:00:37.000
And Alan Polos, who. Please put your hands together for him. He just… we accepted a faculty appointment in Santiago, a Catholic University of Chile! Woo!

00:00:37.000 --> 00:00:47.000
Ah, congratulations! Um, Alan, would you like to introduce yourself first, to follow up on… on that?

00:00:47.000 --> 00:00:48.000
announcement?

00:00:48.000 --> 00:01:00.000
Oh, okay, Cheryl. Yeah, so I'm Alan Poulos, I'm a… postdoc here at, uh… And the Earthquake Science Center at USGS, and uh… Yeah, thank you, Sarah, for those.

00:01:00.000 --> 00:01:03.000
With that, congratulations.

00:01:03.000 --> 00:01:13.000
My name is Cheryl Rubinovich, I'm a researcher and consultant in natural hazard mitigation public policy. I'm based in Northern California, in the Bay Area.

00:01:13.000 --> 00:01:23.000
And I'm very excited to be part of this panel, which broadens out from the physical phenomenon we've been talking about for most of the day so far, into the prediction.

00:01:23.000 --> 00:01:34.000
of consequences and prevention of earthquake impacts, um, in other words, to the social, political, and economic phenomenon that intermesh with physical phenomenon to create.

00:01:34.000 --> 00:01:39.000
the outcomes that we sometimes like, sometimes don't like, sometimes want to change, sometimes can't change.

00:01:39.000 --> 00:01:47.000
Um, so we will again be following the same format, four talks, four unique contributors and topics.

00:01:47.000 --> 00:01:51.000
I'm gonna introduce… we're gonna alternate who's introducing who.

00:01:51.000 --> 00:01:58.000
Um, so to start with, we have David Bonowitz, who's a leading structural engineering practitioner based in.

00:01:58.000 --> 00:02:09.000
San Francisco, and with deep experience into just about everything, including designing solutions, but also advising on code development, city programs.

00:02:09.000 --> 00:02:27.000
um, professional… organizations and how they run, and ultimately policy, so I can't think of anybody better to provide this update on retrofit program progress in the Bay Area, and probably he'll pop in some ideas going beyond Northern California. So here's David Bowitz.

00:02:27.000 --> 00:02:33.000
Good afternoon. Uh, I'm gonna talk about seismic retrofit programs in the Bay Area.

00:02:33.000 --> 00:02:36.000
I'm going to quickly go through some of the programs that are…

00:02:36.000 --> 00:02:41.000
in process right now, talk… and then talk about where those programs came from,

00:02:41.000 --> 00:02:44.000
What's working, maybe what's not.

00:02:44.000 --> 00:02:50.000
And then, on the slide, you can see a list of emerging issues that I'm not gonna get to, but have a look at those, and if any of those

00:02:50.000 --> 00:02:52.000
look interesting, we can talk about them.

00:02:52.000 --> 00:02:54.000
in the Q&A.

00:02:54.000 --> 00:03:00.000
The first thing to know is that retrofits have been going on all over the Bay Area for years.

00:03:00.000 --> 00:03:03.000
what I'm going to talk about in this presentation is the more recent trend

00:03:03.000 --> 00:03:05.000
I would say 10 to 15 years.

00:03:05.000 --> 00:03:10.000
of privately owned buildings being subject to mandatory retrofit programs

00:03:10.000 --> 00:03:19.000
in individual cities. That's relatively new. So in this talk, I'm not going to talk about the public building programs for University of California.

00:03:19.000 --> 00:03:21.000
CSU Public Schools,

00:03:21.000 --> 00:03:25.000
that are regulated by DSA in Sacramento. I'm also not going to talk about

00:03:25.000 --> 00:03:32.000
The electrophits of cripple wall houses, or of unreinforced masonry buildings, which go back 40 years,

00:03:32.000 --> 00:03:36.000
or the state-run program for retrofit of our hospitals.

00:03:36.000 --> 00:03:39.000
Those have been going for a long time.

00:03:39.000 --> 00:03:45.000
I'm gonna use a color coding here to focus on mandatory retrofit programs, as opposed to evaluations

00:03:45.000 --> 00:03:54.000
or background studies, the white markers will indicate incentivized, or some people call them voluntary programs, and I'm also going to distinguish

00:03:54.000 --> 00:04:02.000
the type of building in this way, since most of the programs are premised on a certain structure type or certain structural deficiency.

00:04:02.000 --> 00:04:13.000
We can start with everything except woodsoft story, and if we do that, let's jump in and look in San Francisco has a facade program that is ongoing. It's mostly a maintenance program.

00:04:13.000 --> 00:04:21.000
But it has, obviously, some seismic component to it. San Francisco has a mandatory evaluation program for its private schools.

00:04:21.000 --> 00:04:27.000
And just recently, finally began its program for concrete and tilt-up buildings.

00:04:27.000 --> 00:04:36.000
And this is mandatory, but it's screening only, so it was supposed to be retrofit, and they pulled that back a couple years ago. Over in Oakland,

00:04:36.000 --> 00:04:40.000
They have a nascent idea of a concrete program that was really just

00:04:40.000 --> 00:04:45.000
a city council resolution at the end of Councilman Dan Kalb's term.

00:04:45.000 --> 00:04:53.000
I'm not sure anything's really gonna come of that. I haven't heard anything moving on it. In Berkeley, they've done a number of programs for concrete tilt-up and

00:04:53.000 --> 00:05:03.000
Would commercial buildings, uh, using FEMA grants, but while there's some uptake, it's really, uh, not very effective at making change on a citywide scale.

00:05:03.000 --> 00:05:10.000
And that's gonna be a theme that voluntary or incentivized programs really don't make a change at a citywide scale.

00:05:10.000 --> 00:05:14.000
Down south, Fremont had a tilt-up program.

00:05:14.000 --> 00:05:21.000
Some years ago, I don't know much about it. Over in Palo Alto, they're doing a more exhaustive inventory of their building stocks, thinking about

00:05:21.000 --> 00:05:34.000
various, uh, types of buildings together. The O there includes some steel buildings, some public assembly occupancy buildings, and, uh, even just some buildings

00:05:34.000 --> 00:05:38.000
that have higher than normal occupant load, regardless of their structure type.

00:05:38.000 --> 00:05:43.000
But now let's move to the Woodsoft story buildings, which is the most common type currently going on.

00:05:43.000 --> 00:05:47.000
In the Bay Area. And to be clear, a soft story means

00:05:47.000 --> 00:05:49.000
A wood frame,

00:05:49.000 --> 00:05:56.000
story for multifamily residential building built before, usually 1980, sometimes up to around 1990.

00:05:56.000 --> 00:06:05.000
So here we see three programs that are essentially complete. Fremont had a program. I don't think they captured all the buildings in Fremont.

00:06:05.000 --> 00:06:07.000
That's a question in San Francisco.

00:06:07.000 --> 00:06:16.000
is essentially done with 4,900 retrofits, and over in Berkeley, about 300 retrofits. Again, mandatory programs essentially complete.

00:06:16.000 --> 00:06:24.000
Following that lead, other cities have mandatory retrofit programs in progress in Oakland, in Albany, and in Mill Valley.

00:06:24.000 --> 00:06:35.000
And the rest of the programs around the Bay Area are not mandatory retrofit, but there's some kind of a study, either an ordinance was passed to require only evaluation or screening, but not retrofit.

00:06:35.000 --> 00:06:42.000
or they're in the process of doing their inventory, or at least they're thinking about it, and might eventually get to a retrofit program.

00:06:42.000 --> 00:06:46.000
The white marker down at San Jose, I'll talk about a little bit later.

00:06:46.000 --> 00:06:53.000
I do want to mention that the state has at least the potential for a program, the California Residential Mitigation Program,

00:06:53.000 --> 00:06:56.000
jointly run by CEA and Cal OES.

00:06:56.000 --> 00:07:04.000
Uh, it has the opportunity to get some funding and some funding from the legislature was provided, and then taken away,

00:07:04.000 --> 00:07:07.000
And then they got a FEMA grant, and then it was taken away.

00:07:07.000 --> 00:07:13.000
But the possibilities still exist there. We may talk about that a little bit later, too. So, what

00:07:13.000 --> 00:07:18.000
what are cities thinking about when they think about these soft-story buildings?

00:07:18.000 --> 00:07:28.000
What did we learn from the inventories? So, uh, on the right side of the screen, you can see this is the percentage of housing units citywide

00:07:28.000 --> 00:07:37.000
that happen to be in these collapse-prone, soft-story buildings, if you only think about the buildings with 5 or more units. If you include the buildings with 3 or 4 units,

00:07:37.000 --> 00:07:40.000
Which is common outside of the larger cities,

00:07:40.000 --> 00:07:42.000
You get the column on the left.

00:07:42.000 --> 00:07:48.000
So, down at the bottom, the first cities to do this in the Bay Area with mandatory programs, San Francisco and Oakland,

00:07:48.000 --> 00:07:51.000
We're looking at 14-15% of their housing stock.

00:07:51.000 --> 00:07:58.000
being in these large, bad buildings. So that's a big number. The smaller cities, the more recent programs,

00:07:58.000 --> 00:08:03.000
do not have quite those numbers, but if you add in the 3 and 4 unit buildings,

00:08:03.000 --> 00:08:06.000
They are getting up into the 10-12% range.

00:08:06.000 --> 00:08:11.000
San Jose has smaller numbers, but that's only because San Jose is huge.

00:08:11.000 --> 00:08:20.000
If you look at certain neighborhoods in San Jose, certain zipodes, you'll find that the percentage of housing units in these buildings actually is 10-15%, sometimes even higher.

00:08:20.000 --> 00:08:23.000
So it's really a neighborhood effect…

00:08:23.000 --> 00:08:28.000
There. So, how would you decide as a city that you want a program like this?

00:08:28.000 --> 00:08:31.000
Excuse me. Theoretically…

00:08:31.000 --> 00:08:37.000
You could take all of the buildings in your city, do an inventory, look at them and estimate

00:08:37.000 --> 00:08:46.000
where each one falls on a scale of, for example, the expected recovery time after a design-level earthquake, or you could think about safety, you could think about cost,

00:08:46.000 --> 00:09:00.000
Any metric you like. And if you did an inventory of this, you might see a distribution something like this. Again, this is hypothetical, and if you were to implement a mitigation agenda over the course of a couple decades, you might be able to change that distribution,

00:09:00.000 --> 00:09:13.000
And while that looks like it might not be a big difference, I think it is. That's an important difference, because what you're doing with your agenda is you're reaching into the red and black areas down near the bottom of the left bar,

00:09:13.000 --> 00:09:16.000
and retrofitting those. So, for example, your Soft Story program,

00:09:16.000 --> 00:09:20.000
is looking for collapse-prone buildings that are down in the bottom of that bar,

00:09:20.000 --> 00:09:27.000
Requiring them to be retrofit, so when you're done, they represent a different portion of the building stock.

00:09:27.000 --> 00:09:35.000
And you can do this with other mandatory or voluntary programs for buildings of different types, including for new buildings.

00:09:35.000 --> 00:09:42.000
If you're a city, how do you know which ones are most critical? Where would you start? One way to think about that

00:09:42.000 --> 00:09:47.000
Engineers think about the different structural deficiencies, and you can see some of them listed across the top.

00:09:47.000 --> 00:09:55.000
of this matrix, but for community resilience purposes, we also want to think about the use and function of these buildings. So there's a partial list

00:09:55.000 --> 00:10:02.000
down the side. And traditionally, we've been working at the edges of this matrix, for example, unreinforced masonry buildings.

00:10:02.000 --> 00:10:06.000
Regardless of what use they, uh, they serve,

00:10:06.000 --> 00:10:11.000
Those are killer buildings, and they kill people on the sidewalk, so we've been working on these for 40 years.

00:10:11.000 --> 00:10:14.000
Similarly, hospitals are urgent

00:10:14.000 --> 00:10:21.000
Regardless of what the structural deficiency might be, so we have a program for those as well. But those are just the edges.

00:10:21.000 --> 00:10:31.000
In a city, you may find that you have a lot of combinations that are not at the edges, and if you were to do a careful inventory, you might be able to identify certain combinations that will pop out.

00:10:31.000 --> 00:10:37.000
as being highly critical, and this is where cities like San Francisco and Oakland

00:10:37.000 --> 00:10:41.000
basically learning the lessons from Loma Prieta and Northridge,

00:10:41.000 --> 00:10:47.000
Said soft story, multifamily residential buildings are a problem, and we're going to address them through public policy.

00:10:47.000 --> 00:10:49.000
with some kind of a mitigation program.

00:10:49.000 --> 00:11:01.000
Now, is this the kind of analysis that every city does? No. In fact, even San Francisco and Oakland did not do this kind of inventory, they just kind of knew what their inventory was to begin with.

00:11:01.000 --> 00:11:06.000
But other cities wouldn't do this in the first place because they are smartly,

00:11:06.000 --> 00:11:10.000
Not trying to be leaders. They don't have to figure this out for the first time.

00:11:10.000 --> 00:11:15.000
If San Francisco is willing to figure out all the logistics and set the technical criteria,

00:11:15.000 --> 00:11:18.000
and create a market of contractors and engineers,

00:11:18.000 --> 00:11:22.000
And experiment with the right legislative language,

00:11:22.000 --> 00:11:24.000
Why not just follow their lead?

00:11:24.000 --> 00:11:27.000
I'm all for that, I think it's a smart way for a smaller city

00:11:27.000 --> 00:11:34.000
to go. If you're a city, whether a leader or a follower, and you want to have a program like this, you have a couple of decisions.

00:11:34.000 --> 00:11:39.000
To make, one of them is what kind of work you're going to require, and here we have a…

00:11:39.000 --> 00:11:46.000
a list going from a low-effort, just send notices to owners, to a very high effort, a much more complicated retrofit.

00:11:46.000 --> 00:11:50.000
The second choice you have to make is whether this program is gonna be

00:11:50.000 --> 00:11:52.000
incentivized or mandatory.

00:11:52.000 --> 00:12:00.000
So, there, I want to call your attention to, on this combination for target story structural retrofit has turned out to

00:12:00.000 --> 00:12:07.000
to be the consensus, best bang for your buck, just focus on the collapse-prone condition,

00:12:07.000 --> 00:12:09.000
Don't worry about some of the other things.

00:12:09.000 --> 00:12:13.000
This is a very effective way to do it, and in particular, go mandatory.

00:12:13.000 --> 00:12:22.000
Because that's a much more effective, uh, risk reduction at a citywide scale. Now, 10 to 15 years ago, when these programs got started, only Fremont

00:12:22.000 --> 00:12:32.000
was willing to go into the yellow box. Other cities were thinking about the same risk, but, you know, Oakland, Berkeley, they said, yes, we want to be mandatory, but we're not going to make you do the retrofit.

00:12:32.000 --> 00:12:41.000
So we're gonna stop with screening, we're gonna stop with evaluation. Maybe if you find out that you've got a bad building, you'll voluntarily retrofit.

00:12:41.000 --> 00:12:43.000
San Francisco, for example,

00:12:43.000 --> 00:12:51.000
wanted to say, well, we want to focus you on retrofit, but we're not going to make you do it, we're just going to try to provide some incentive to try to nudge you

00:12:51.000 --> 00:12:53.000
into doing this on your own.

00:12:53.000 --> 00:13:00.000
Neither approach worked. So they didn't get a lot of uptake from anybody to actually do retrofit to reduce the real risk.

00:13:00.000 --> 00:13:05.000
But that was early on. What happened after they learned that this wasn't working is that they…

00:13:05.000 --> 00:13:10.000
Most of these cities eventually moved to mandatory retrofit,

00:13:10.000 --> 00:13:19.000
and San Francisco and Berkeley put themselves in the yellow box. Oakland went even further and said, we're gonna do that. Plus, we're gonna ask for a little bit of non-structural stuff as well.

00:13:19.000 --> 00:13:30.000
And by leading the way like that, it made it easier for Albany and Mill Valley and San Jose to follow, and that's why they have mandatory retrofit programs as well.

00:13:30.000 --> 00:13:41.000
Except for San Jose. So, San Jose actually passed an ordinance for a mandatory retrofit program, and it is still in their municipal code, but it looks like they're gonna back off of that.

00:13:41.000 --> 00:13:48.000
Uh, and move to an incentivized program, at least to start with. Why? Because they have

00:13:48.000 --> 00:13:57.000
Why? Because I think they have, in my mind, a misunderstanding of the financing issues. So now we have to talk about the money a little bit.

00:13:57.000 --> 00:14:05.000
A soft story retrofit ballpark, think of $100,000 per building, but obviously there's a wide range because the buildings vary in size.

00:14:05.000 --> 00:14:13.000
Uh, complexity, etc. A typical FEMA grant is on the order of $4-5 million, so if you're providing a partial subsidy,

00:14:13.000 --> 00:14:21.000
for these buildings, if you've got 200 buildings, yeah, you can make a real big dent with a $4-5 million dollar grant.

00:14:21.000 --> 00:14:25.000
If you've got 3,000 buildings in your program, like San Jose,

00:14:25.000 --> 00:14:27.000
then you're only going to be able to help

00:14:27.000 --> 00:14:32.000
100 to 150 buildings, it's a very small percentage of your overall citywide program.

00:14:32.000 --> 00:14:37.000
So, San Jose looked at that and said, we need to provide assistance for everybody else. They're wrong.

00:14:37.000 --> 00:14:49.000
But that's what they thought, and therefore, because they couldn't get traction on their FEMA program, because FEMA's in disarray, and they couldn't get a second grant, and they couldn't find the money in the internal city budget that they

00:14:49.000 --> 00:14:52.000
Housing Director had promised to find,

00:14:52.000 --> 00:14:55.000
They have backed off of their mandatory program,

00:14:55.000 --> 00:14:58.000
And are now looking at a smaller incentivized program.

00:14:58.000 --> 00:15:00.000
At least for now.

00:15:00.000 --> 00:15:06.000
So, to talk about the FEMA grant programs, uh, as you know, FEMA is in flux, I would even say in crisis,

00:15:06.000 --> 00:15:09.000
And just for purposes of discussion,

00:15:09.000 --> 00:15:15.000
I throw out the idea that maybe the Western states could have a coalition where they don't have to rely on FEMA,

00:15:15.000 --> 00:15:26.000
and can find some way to fund their own mitigation programs. But the real question is, to do this kind of work, soft story mitigation, do you really need a grant at all?

00:15:26.000 --> 00:15:33.000
San Francisco did 4,900 buildings with no FEMA money at all. Los Angeles did 14,000 buildings.

00:15:33.000 --> 00:15:42.000
With no FEMA money at all. And yet, San Jose was looking at this and said, well, we need to provide assistance, or we can't get political feasibility,

00:15:42.000 --> 00:15:45.000
To go forward with this.

00:15:45.000 --> 00:15:50.000
And, you know, they could be looking at San Francisco and Oakland and saying, well, they tried and it failed.

00:15:50.000 --> 00:15:55.000
So why are we not learning the lesson from them? And I think it goes back to this question

00:15:55.000 --> 00:16:02.000
of leaders and followers, and whether you are looking for good examples when you start creating your program.

00:16:02.000 --> 00:16:09.000
If you're a follower, you're gonna look for those precedents, and you'll quickly learn that San Francisco and Oakland went down that road, and it didn't really work.

00:16:09.000 --> 00:16:15.000
And that by going to a mandatory rectifit program, they have now demonstrated that it's completely feasible.

00:16:15.000 --> 00:16:21.000
for owners to do this work without assistance. And yet, uh, that's what you would learn if you were a follower.

00:16:21.000 --> 00:16:24.000
But if you're a follower, but you think you're a leader,

00:16:24.000 --> 00:16:31.000
That's a bad combination, and you may find that you're not gonna go looking for precedence and examples to follow, and you…

00:16:31.000 --> 00:16:35.000
may fall into the trap of convincing yourself that you need to find money.

00:16:35.000 --> 00:16:37.000
Before you can go forward with a mandatory program.

00:16:37.000 --> 00:16:42.000
To which I say that if you've got a real seismic risk, it's going to affect your housing and your community,

00:16:42.000 --> 00:16:46.000
But it's only important enough to address if somebody else will pay for it.

00:16:46.000 --> 00:16:49.000
then you don't really believe in your risk.

00:16:49.000 --> 00:16:57.000
Anyway, that's… we're getting into my personal opinions there, so I'll leave it at that, and we can discuss it more, uh, in the Q&A.

00:16:57.000 --> 00:17:03.000
Again, there's a list of other issues. Anything you want to ask is fine. These issues, or anything I talked about, or anything else.

00:17:03.000 --> 00:17:12.000
Thank you very much.

00:17:12.000 --> 00:17:20.000
Thank you, David. So, we'll continue with our… Next speaker, who's, uh, John van der Lindt.

00:17:20.000 --> 00:17:27.000
from Colorado State University. And John's research focuses on performance-based engineering.

00:17:27.000 --> 00:17:32.000
We have applications and many. Uh, natural hazards.

00:17:32.000 --> 00:17:37.000
And, uh, John will talk about estimating the cost for earthquakes to vulnerable.

00:17:37.000 --> 00:17:41.000
communities using the in-core software.

00:17:41.000 --> 00:17:51.000
like to acknowledge the project collaborators, two former postdocs at Colorado State University, as well as, um,

00:17:51.000 --> 00:17:54.000
Harvey Cutler, Professor Emeritus in economics.

00:17:54.000 --> 00:18:07.000
Uh, and colleagues at the California Seismic Safety Commission, Jawan Connolly and Sarah McBride.

00:18:07.000 --> 00:18:16.000
vulnerable groups, including low-income households, racial and ethnic minorities, and individuals with limited access to education and healthcare,

00:18:16.000 --> 00:18:24.000
Um, there's a significant disparities in disaster recovery, primarily the speed of recovery.

00:18:24.000 --> 00:18:39.000
The approach presented today uses NCORE, and it models the interdependencies between physical aspects of a community, and this is essentially the earthquake-induced damage, which results in a loss of functionality of buildings.

00:18:39.000 --> 00:18:47.000
And then looking at the economic system model, something called a dynamic computable general equilibrium model.

00:18:47.000 --> 00:18:55.000
And we can look at employment shifts, domestic supply by sector, as well as the household income distribution.

00:18:55.000 --> 00:19:04.000
And we're able to then model the speed of recovery of a community, the economic metrics themselves.

00:19:04.000 --> 00:19:17.000
So, in CORE, the, uh, the interdependent Network Community Resilience Modeling environment, within that, the goal is to be able to predict damage probabilities at the individual building level.

00:19:17.000 --> 00:19:25.000
Using physics so that it's transferable anywhere. And we do this using fragilities.

00:19:25.000 --> 00:19:36.000
also to predict not just how many, but which households will dislocate based on their race, ethnicity, their income, other characteristics. And we do this by synthetically allocating

00:19:36.000 --> 00:19:38.000
Uh, uh…

00:19:38.000 --> 00:19:49.000
people or families at the household level. We do this by deaggregating census block-level data. And if we aggregate that back up, then we would reproduce the census at the block level.

00:19:49.000 --> 00:19:57.000
And this is done with something called our Housing Unit Allocation.

00:19:57.000 --> 00:20:11.000
than to predict changes in community-level incomes by household demographic, changes in gross product, and changes in the tax base by sector. And we do this with that computable general equilibrium model that I mentioned.

00:20:11.000 --> 00:20:17.000
And then have the information to make resilience-informed, socially equitable decisions.

00:20:17.000 --> 00:20:20.000
Uh, and that is done within CORE.

00:20:20.000 --> 00:20:26.000
In community resilience, we look at four areas of community stability.

00:20:26.000 --> 00:20:32.000
Within the parentheses, you'll see a number, or at least several different metrics. Of course, there's a number of others.

00:20:32.000 --> 00:20:34.000
Um, and this allows us to…

00:20:34.000 --> 00:20:38.000
Uh, to measure this, but these are populations stability,

00:20:38.000 --> 00:20:46.000
economic stability, social services stability, and physical services stability.

00:20:46.000 --> 00:20:52.000
So, the modeling of a community in core modeling begins with an integrated community model.

00:20:52.000 --> 00:21:02.000
Uh, and that is using a suite of building archetypes that are assigned to all the building in the community, and so we're looking at the individual building level.

00:21:02.000 --> 00:21:05.000
can model the transportation network, the water network,

00:21:05.000 --> 00:21:14.000
the electric power network, as well as those dependencies, uh, between networks and buildings and the networks themselves.

00:21:14.000 --> 00:21:24.000
On the top right, this is a representation of the computable general equilibrium model, and this is just showing, sort of, that conceptual representation.

00:21:24.000 --> 00:21:32.000
And in supporting field studies to develop the data-driven models that you see on the bottom right. That was a study in 2016.

00:21:32.000 --> 00:21:38.000
of Lumberton, North Carolina, that went all the way to 2024.

00:21:38.000 --> 00:21:40.000
a longitudinal field study.

00:21:40.000 --> 00:21:47.000
And this is to model households as well as social institutions.

00:21:47.000 --> 00:21:56.000
So the interdependent Network Community Resilience Modeling environment, we're able to look at, once we have that integrated community model, that's the

00:21:56.000 --> 00:22:02.000
the initial condition, we're able to look at the physical infrastructure, the economic health, the social services,

00:22:02.000 --> 00:22:07.000
And this is all done, uh, within a cloud computing system.

00:22:07.000 --> 00:22:16.000
with our partners at the National Center for Supercomputing Application, and that's free to anyone to use and available

00:22:16.000 --> 00:22:20.000
Uh, available just simply by signing up.

00:22:20.000 --> 00:22:25.000
So once we have that, we then this description of the community, we then

00:22:25.000 --> 00:22:32.000
can model the damage and the loss, the impact to natural hazards, in this case, to an earthquake.

00:22:32.000 --> 00:22:38.000
And then look at the multidisciplinary recovery with these integrated databases.

00:22:38.000 --> 00:22:43.000
And the whole idea of NCORE is to look at alternative actions to enhance community resilience.

00:22:43.000 --> 00:22:50.000
to inform planning, effectively performing what-if scenarios.

00:22:50.000 --> 00:22:56.000
I won't go through this in detail, obviously, but this is, uh, this description that I mentioned.

00:22:56.000 --> 00:23:00.000
Uh, and then we can look at our damage here.

00:23:00.000 --> 00:23:03.000
Uh, using fragilities and using a hazard model.

00:23:03.000 --> 00:23:07.000
Uh, in this case, it would be a raster of peak

00:23:07.000 --> 00:23:18.000
peak ground accelerations or spectral accelerations, velocities, displacements, depending on what we're looking at.

00:23:18.000 --> 00:23:25.000
And then we can look at the state of the recovery over time, and you can see this looping through time, or at least time steps.

00:23:25.000 --> 00:23:34.000
And the idea is then updating that CGE model to then see how fast a community recovers.

00:23:34.000 --> 00:23:44.000
And then we can look at the policy levers and the decision combinations, and in this case, we're looking at, uh, FEMA assistance.

00:23:44.000 --> 00:24:01.000
So I mentioned the economic model is key to this analysis. This is a spatial dynamic computable general equilibrium model. In the case of these small communities, it's not as much spatial, simply because their populations are so small.

00:24:01.000 --> 00:24:09.000
but we're able to incorporate infrastructure damage and population disruptions, um, provide a range of economic and fiscal impacts.

00:24:09.000 --> 00:24:20.000
code chain… look at code changes, um, and… and then we can look at losses in production, wages, jobs in the local sector.

00:24:20.000 --> 00:24:38.000
The social science models, which we're utilizing a little bit here, but not quite as much. We can look at population and employee dislocation, housing restoration and recovery, which is important, and of course, business interruption and restoration.

00:24:38.000 --> 00:24:53.000
So, this is just an example of, uh, of a web-based app that runs in Core, and in this case, it's a recording, but this is something that we use with our partner, Salt Lake City.

00:24:53.000 --> 00:24:58.000
Uh, and this was a series of workshops to enable them to,

00:24:58.000 --> 00:25:13.000
work on their Fix the Bricks program, which was a retrofit of unreinforced masonry. So this is linked with the community resilience Planning Guide, that six-step guide that you may or may not have heard of, but this is being used

00:25:13.000 --> 00:25:23.000
around the country by a number of communities in partnership with NIST and others. And so we can take our… we can set up our earthquake. Uh, here you see this is using NGAs.

00:25:23.000 --> 00:25:25.000
And, um…

00:25:25.000 --> 00:25:32.000
pick our attenuation model in this case.

00:25:32.000 --> 00:25:39.000
And then can pick an epicenter, uh, here, just sort of done a little bit randomly, but just to demonstrate.

00:25:39.000 --> 00:25:48.000
Uh, and then this will generate our raster. In this case, I believe it's PGA, I can't remember what we decided, but…

00:25:48.000 --> 00:25:57.000
And I'll jump ahead just a little, uh, and so now, um, we're going to look at

00:25:57.000 --> 00:26:05.000
We're gonna look at the damage to the buildings. Here you see the performance, and this is what we're trying to get with this particular

00:26:05.000 --> 00:26:09.000
earthquake, we can look at whether the buildings are functional or non-functional.

00:26:09.000 --> 00:26:17.000
Uh, the types of buildings, these are the building clusters that are in the NIST Community Resilience Planning Guide I mentioned.

00:26:17.000 --> 00:26:27.000
And then we can jump ahead and look at the number of households that are dislocated, and we can do this by tenure, whether they're renters or whether they're owners.

00:26:27.000 --> 00:26:30.000
And then we're able to…

00:26:30.000 --> 00:26:34.000
go over, and of course, this is all… this is all done, uh,

00:26:34.000 --> 00:26:38.000
running in core on its, uh, its server cluster.

00:26:38.000 --> 00:26:45.000
And we could look at, uh, dislocation by household income, race, ethnicity, housing type, things like that as well.

00:26:45.000 --> 00:26:52.000
And then we can get into the changes in the economy, and that's really what we're going to talk about a little bit today, so…

00:26:52.000 --> 00:26:56.000
Um, so with that, I'll go ahead and jump forward.

00:26:56.000 --> 00:27:05.000
So, as you know, uh, as you know, there was a series of two earthquakes in Rio del, California. This is a…

00:27:05.000 --> 00:27:12.000
population, only about 3,400 people, uh, with the majority being out commuters for work.

00:27:12.000 --> 00:27:21.000
And a very low median household income for Northern California.

00:27:21.000 --> 00:27:39.000
we looked for two other communities that had similar characteristics, but that had different levels of income. And so what we see here is East Richmond Heights, which has, uh, approximately the same population, uh, mostly, you know, 9, 9.5 out of 10

00:27:39.000 --> 00:27:46.000
out commuters, but with a median income of $112,000, so effectively our middle income.

00:27:46.000 --> 00:28:02.000
And then Monte Sereno, uh, and this had approximately the same population, same number of route commuters, but with a much higher median household income, an affluent or wealthy community.

00:28:02.000 --> 00:28:06.000
the California Seismic Safety Commission, uh,

00:28:06.000 --> 00:28:13.000
did damage inspection following the Rio del earthquakes, and what we see here is our…

00:28:13.000 --> 00:28:25.000
the damage from those post-earthquake inspections, and knowing which buildings. And so, what we did is we took that level of damage, and then looking at, uh,

00:28:25.000 --> 00:28:36.000
which types of… which archetypes, or which buildings, and whichctors would be effectively modeled by the…

00:28:36.000 --> 00:28:39.000
computable general equilibrium model, we then

00:28:39.000 --> 00:28:51.000
went through and superimposed or reproduced that damage level across the same building types in these other two communities. And so that is… the idea was to compare apples to apples.

00:28:51.000 --> 00:29:01.000
So, in looking at Rio Del, East Richmond Heights, as well as Monte Sereno, what we see is we can see the number of households, and you can see the

00:29:01.000 --> 00:29:12.000
the income disparities, uh, you know, a quarter of… a quarter of the households making less than $30,000, and only a third making even more than $90,000 in Rio Dell.

00:29:12.000 --> 00:29:20.000
Whereas when we get into East Richmond Heights, um, then very few making less than 30,000, it's not marked here, but…

00:29:20.000 --> 00:29:25.000
And then we can see that incomes, we have about a quarter making more than $200,000.

00:29:25.000 --> 00:29:35.000
When we get to the most affluent community in the study, we can see that we have about a third making more than $350,000, and…

00:29:35.000 --> 00:29:39.000
And really, only a third making even less than $130,000.

00:29:39.000 --> 00:29:47.000
in the capital stock, uh, similar, uh, similar situation in that, um, you know, many of the buildings, uh,

00:29:47.000 --> 00:29:53.000
you know, the majority are under $300,000, almost all

00:29:53.000 --> 00:30:02.000
residential structures, and then in East Richmond Heights, we start to see that there's more capital, $108 million worth.

00:30:02.000 --> 00:30:07.000
Uh, for house, uh, housing structures, over 300,000.

00:30:07.000 --> 00:30:17.000
And Monte Sereno, a number, about 222 million for houses over $500,000, and this is consistent with the incomes as we'd expect.

00:30:17.000 --> 00:30:28.000
Uh, the losses from this analysis, and so I should mention that, as you can see within CORE, we're not running the hazard like I showed you for the Salt Lake City example, but we're actually jumping past that.

00:30:28.000 --> 00:30:32.000
And starting immediately at the damage to then run the…

00:30:32.000 --> 00:30:37.000
the simulation. And so what you see here is, um, is that…

00:30:37.000 --> 00:30:51.000
Uh, in this, I think this total shows most of it, is that the capital stock losses of $21 million is about 14%. Here, 13% is $49 million. And here, 13% is almost $80 million, and that's showing that…

00:30:51.000 --> 00:30:57.000
the difference in capital, including the loss.

00:30:57.000 --> 00:31:01.000
And then we can look at their utilized savings. This savings, uh…

00:31:01.000 --> 00:31:08.000
and looking at this, uh, in the ratio of capital stock to utilize savings, that ratio.

00:31:08.000 --> 00:31:16.000
And you can see that depending on income, of course, this goes up, and that the highest income bracket in Monte Sereno has

00:31:16.000 --> 00:31:22.000
almost twice as much savings as capital stock loss.

00:31:22.000 --> 00:31:25.000
So what's actually happening? Well…

00:31:25.000 --> 00:31:36.000
Um, with no financial assistance from FEMA, or in general, then what we see here is in poor communities, we have lower spending power, the businesses struggle,

00:31:36.000 --> 00:31:47.000
fewer opportunities for reskilling or upskilling, job losses, uh, in the infrastructure, you know, there's the lower disposable income and financial constraints, slower recovery.

00:31:47.000 --> 00:31:53.000
of the domestic supply of goods. In the wealthier community, or a wealthy community,

00:31:53.000 --> 00:32:04.000
There's higher disposable income supporting local businesses during recovery, higher, uh, more specialized skills, higher consumer demand and spending power.

00:32:04.000 --> 00:32:08.000
Um, and so, uh, businesses can quickly

00:32:08.000 --> 00:32:13.000
restore their supply chain.

00:32:13.000 --> 00:32:28.000
So, the financial assistance from FEMA, if we look at this, this would be our… we have our number of damaged housing units within the different brackets. We have our capital stock losses in millions.

00:32:28.000 --> 00:32:35.000
Uh, and then we have the FEMA funds that would have been injected. Now, they weren't in this case, as you know, or may know.

00:32:35.000 --> 00:32:42.000
And what we see is the ratio of FEMA funds to the capital stock loss, and you can see it's much more significant, as you would expect.

00:32:42.000 --> 00:32:49.000
Uh, for the poor, uh, poorer community, uh, than the wealthier community.

00:32:49.000 --> 00:32:55.000
So what does this mean? Well, the recovery periods for real household income.

00:32:55.000 --> 00:33:04.000
Uh, through the model. And so these are quarters, these are approximately 3-month periods, and what we see is we see that

00:33:04.000 --> 00:33:09.000
Uh, that with no assistance, um, the wealthier or the slightly

00:33:09.000 --> 00:33:12.000
wealthier, uh…

00:33:12.000 --> 00:33:17.000
households, even within the same community, recover quicker, and this is consistent everywhere.

00:33:17.000 --> 00:33:26.000
With no assistance. With delayed assistance, which is assistance after 2 years, we can see that the significantly improves

00:33:26.000 --> 00:33:28.000
for the poorer households,

00:33:28.000 --> 00:33:33.000
Less so for when we're getting into, say, middle-class household.

00:33:33.000 --> 00:33:36.000
And then… or middle-income household.

00:33:36.000 --> 00:33:43.000
And then for immediate assistance, which was at 6 months, we can see this improves, but not that much over…

00:33:43.000 --> 00:33:46.000
this, and in fact, uh,

00:33:46.000 --> 00:33:50.000
the amount of income, household income,

00:33:50.000 --> 00:33:52.000
doesn't matter as much.

00:33:52.000 --> 00:34:00.000
Again, then looking at East Richmond Heights, our middle-income community, we can see that this disparity

00:34:00.000 --> 00:34:13.000
is also notable, and then goes down, but again, fairly consistent across as we get into delayed assistance, but we can see there's no effect whether the assistance is immediate or delayed.

00:34:13.000 --> 00:34:19.000
And then when we get to our most affluent community, Monte Sereno, we can see that, um,

00:34:19.000 --> 00:34:22.000
That we, uh, uh…

00:34:22.000 --> 00:34:26.000
this has, uh, um…

00:34:26.000 --> 00:34:35.000
recovers fairly quick, and we can see that whether or not they get assistance, uh, seems to not really matter.

00:34:35.000 --> 00:34:50.000
So, the conclusions, and these are supported in the report and in the upcoming paper as well in more detail than a short presentation, but FEMA relies on a state's total taxable resources, which results in a lack of equitable support

00:34:50.000 --> 00:34:52.000
for rural communities within large states.

00:34:52.000 --> 00:35:03.000
And so this study shows that rural, underserved communities like Rio Del will have disproportionately longer recovery times compared with more affluent communities.

00:35:03.000 --> 00:35:16.000
Low-income households within the rural community have the longest recovery period, as you saw. And that FEMA impact… excuse me, FEMA assistance has more impact on these poorer communities, as you saw from that table.

00:35:16.000 --> 00:35:23.000
And really almost no impact on the speed of recovery for the wealthier communities.

00:35:23.000 --> 00:35:43.000
So, some recommendations were direct cash transfers, uh, get the money into the hands of the low-income households as quick as possible. Low-interest loans, I know there's some programs out there already, uh, but looking at specifically aiding repair and construction of damaged properties, and then individual assistance criteria.

00:35:43.000 --> 00:35:48.000
looking at the totality of the circumstances rather than the state itself.

00:35:48.000 --> 00:36:01.000
And then low-income prioritization, uh, disaster aids, such as individual assistance, should prioritize low-income households, uh, specifically, since they take the longest to recovery.

00:36:01.000 --> 00:36:07.000
And with that, um, I'll just say the report was prepared by Colorado State University, uh, for

00:36:07.000 --> 00:36:15.000
the California Seismic Safety Commission, uh, within the governor's, uh, Office of Emergency Services, and, uh,

00:36:15.000 --> 00:36:18.000
We're grateful, um, grateful to…

00:36:18.000 --> 00:36:34.000
Uh, for that assistance, and with that, um, thank you.

00:36:34.000 --> 00:36:38.000
Cheryl, you're muted.

00:36:38.000 --> 00:36:49.000
Beautiful. Thank you, John, for that. I would like to now introduce our third speaker, Dr. Saleem Gane is a scholar and researcher in earthquake engineering at.

00:36:49.000 --> 00:36:55.000
UC Berkeley at the Pacific Earthquake Engineering Research Center here.

00:36:55.000 --> 00:36:59.000
Um, he is about to describe the methodology developed to estimate.

00:36:59.000 --> 00:37:08.000
spatially specific benefits of earthquake early warning. Uh, for the case of avoiding passenger train derailment.

00:37:08.000 --> 00:37:17.000
Using BART as an example. Um, that applies the USURF3 catalog through a custom design algorithm.

00:37:17.000 --> 00:37:23.000
This is part of a project for the California Office of Emergency Services that I was also a part of.

00:37:23.000 --> 00:37:28.000
Um, which overall explored the… costs and benefits of EEW.

00:37:28.000 --> 00:37:34.000
More broadly across the state, using both qualitative interviews with EEW users and vendors.

00:37:34.000 --> 00:37:45.000
And additional benefit cost analyses of some other. use cases and… uh, different cost-benefit method comparison.

00:37:45.000 --> 00:37:53.000
So the reports from that study are in review. This is… consider this a teaser of one of the more technical elements that.

00:37:53.000 --> 00:38:02.000
physical science and engineering-oriented elements of the project. Um, and we're happy to answer any questions at the end. Take it away, Salem.

00:38:02.000 --> 00:38:11.000
Good afternoon. My name is Sirin Gunai, and I'm a research and development engineer at the Pacific Earthquake Engineering Research Center Pierre.

00:38:11.000 --> 00:38:14.000
At the University of California, Berkeley.

00:38:14.000 --> 00:38:18.000
In this presentation, I will discuss earthquake early warning benefits,

00:38:18.000 --> 00:38:23.000
And how these benefits, particularly in terms of monetary value,

00:38:23.000 --> 00:38:26.000
are quantified using the rupture scenarios.

00:38:26.000 --> 00:38:30.000
In the 3rd California earthquake Rapture Forecast, or USURF3.

00:38:30.000 --> 00:38:34.000
The project that forms the basis of this work was funded

00:38:34.000 --> 00:38:39.000
by the California Governor's Office of Emergency Services, Cal OES.

00:38:39.000 --> 00:38:43.000
I also would like to acknowledge several colleagues listed here,

00:38:43.000 --> 00:38:47.000
for their contributions and feedback to the project.

00:38:47.000 --> 00:38:50.000
So, this presentation is two primary objectives.

00:38:50.000 --> 00:38:56.000
The first is to explain and demonstrate how the usurf three scenarios are used to…

00:38:56.000 --> 00:38:59.000
provide a holistic assessment of EEW,

00:38:59.000 --> 00:39:02.000
with a particular focus on benefits.

00:39:02.000 --> 00:39:08.000
The second objective is to discuss the regional scale spatial outcomes,

00:39:08.000 --> 00:39:13.000
and aggregated results that can be obtained using this methodology.

00:39:13.000 --> 00:39:25.000
So, as we know, the USGS ShakeAlert EW system operates in the West Coast states and detects earthquakes using P wave arrivals before strong shaking occurs.

00:39:25.000 --> 00:39:28.000
And provide seconds of warning.

00:39:28.000 --> 00:39:31.000
And as shown in this graphic from this website,

00:39:31.000 --> 00:39:38.000
Uh, when an earthquake occurs, two types of seismic waves propagate from the source.

00:39:38.000 --> 00:39:42.000
The faster propagating P waves, which are generally weaker,

00:39:42.000 --> 00:39:45.000
And do not produce strong shaking.

00:39:45.000 --> 00:39:50.000
And the slower S waves that result in the actual strong motion.

00:39:50.000 --> 00:39:55.000
So, using the arrival time of the P waves to a large array of seismograms,

00:39:55.000 --> 00:40:00.000
The ShakeAlert system estimates the location and magnitude of an earthquake,

00:40:00.000 --> 00:40:02.000
And provides alerts.

00:40:02.000 --> 00:40:06.000
So this EW notification is used for several purposes.

00:40:06.000 --> 00:40:10.000
Including smartphone notifications to alert

00:40:10.000 --> 00:40:14.000
public, and for several automated actions, such as

00:40:14.000 --> 00:40:19.000
Slowing or stopping trains from mass transit systems?

00:40:19.000 --> 00:40:21.000
Chatting of gas valves,

00:40:21.000 --> 00:40:24.000
Moving elevators to the closest floor, and…

00:40:24.000 --> 00:40:27.000
Opening the bay doors of fire stations.

00:40:27.000 --> 00:40:31.000
Here, I will, uh, focus on the mass transit mainly.

00:40:31.000 --> 00:40:39.000
is the methodology that we developed for the mass transit covers all essential elements of EEW.

00:40:39.000 --> 00:40:41.000
Including shaking intensity,

00:40:41.000 --> 00:40:45.000
The infrastructure response and warning time.

00:40:45.000 --> 00:40:49.000
So, as a mass transit system in Northern California,

00:40:49.000 --> 00:40:53.000
Particularly, we focused on the Bay Area Rapid Transit.

00:40:53.000 --> 00:41:01.000
BART, and computed the EEW monetary benefits for approximately 185,000.

00:41:01.000 --> 00:41:04.000
Rapture scenarios in USERF3.

00:41:04.000 --> 00:41:07.000
So, how did we calculate these monetary benefits?

00:41:07.000 --> 00:41:12.000
The methodology is based on discretizing the mass transit system

00:41:12.000 --> 00:41:15.000
into what we call a set of points of interest.

00:41:15.000 --> 00:41:20.000
computing the EEW benefit in terms of monetary value at each point.

00:41:20.000 --> 00:41:23.000
And summing the benefits across all points.

00:41:23.000 --> 00:41:26.000
So, here we see a map of the BART system.

00:41:26.000 --> 00:41:31.000
And we divided into one-mile trek segments,

00:41:31.000 --> 00:41:37.000
And the midpoint of each segment shown with circles serves as a point of interest.

00:41:37.000 --> 00:41:42.000
So, before explaining how we calculate the EEW benefit at each point,

00:41:42.000 --> 00:41:46.000
It is insightful to look into…

00:41:46.000 --> 00:41:52.000
How EW can reduce the harm and cause benefits in the mass transit use case.

00:41:52.000 --> 00:41:59.000
So, a particular benefit of EEW is to reduce the probability of derailment

00:41:59.000 --> 00:42:02.000
By automatic de-slowing or stopping trains,

00:42:02.000 --> 00:42:04.000
And actually, the relevant of trains…

00:42:04.000 --> 00:42:08.000
is a common phenomenon observed in earthquakes,

00:42:08.000 --> 00:42:11.000
And the element of trains were observed.

00:42:11.000 --> 00:42:16.000
Uh, drainment of trains were prevented in several earthquakes,

00:42:16.000 --> 00:42:19.000
In Japan and elsewhere as a result of…

00:42:19.000 --> 00:42:24.000
you know, slowing and stopping trains upon receiving an EWL.

00:42:24.000 --> 00:42:32.000
As I mentioned, EW reduces the train derailment probability by automated stopping or slowing down of trains,

00:42:32.000 --> 00:42:39.000
Therefore, our starting point is the calculation of the derailment probability reduction due to EEW.

00:42:39.000 --> 00:42:43.000
We use the fragility curve for this purpose that you see here.

00:42:43.000 --> 00:42:48.000
Although such fragility curves are not available commonly for trains,

00:42:48.000 --> 00:42:50.000
We adopt the fragility curve

00:42:50.000 --> 00:42:58.000
from literature, with the median and dispersion parameters broadly consistent with the seismic response of BART trains.

00:42:58.000 --> 00:43:02.000
The x-axis is the intensity of shaking,

00:43:02.000 --> 00:43:06.000
Expressed here in terms of the peak ground acceleration, PGA.

00:43:06.000 --> 00:43:12.000
While other relevant intensity measures, such as the peak ground velocity, are also possible.

00:43:12.000 --> 00:43:15.000
And the x-axis on the top is MMI.

00:43:15.000 --> 00:43:18.000
modified Merkeley intensity.

00:43:18.000 --> 00:43:22.000
Which is the intensity measure used by BART in their EW system.

00:43:22.000 --> 00:43:26.000
And which is calculated as a function of PGA,

00:43:26.000 --> 00:43:29.000
Using an empirical equation.

00:43:29.000 --> 00:43:32.000
Y-axis is the probability of derailment.

00:43:32.000 --> 00:43:38.000
The probability of derailment depends on the intensity of shaking and the train speed.

00:43:38.000 --> 00:43:45.000
The dependency on the intensity of shaking is obvious, but why the dependency on the train speed?

00:43:45.000 --> 00:43:48.000
There are mainly two reasons for

00:43:48.000 --> 00:43:51.000
This, first, is that…

00:43:51.000 --> 00:43:54.000
Train tracks contain… contain imperfections.

00:43:54.000 --> 00:43:58.000
And when a train travels at regular speed,

00:43:58.000 --> 00:44:03.000
These imperfections can cause real-rail contact irregularities.

00:44:03.000 --> 00:44:06.000
And can introduce lateral accelerations.

00:44:06.000 --> 00:44:12.000
Second reason is the centripetal acceleration caused by

00:44:12.000 --> 00:44:16.000
Train speed on curved tracks.

00:44:16.000 --> 00:44:19.000
So, going back to the intensity of shaking…

00:44:19.000 --> 00:44:23.000
For a given, you serve 3 fault rupture,

00:44:23.000 --> 00:44:26.000
An example of which we see here…

00:44:26.000 --> 00:44:30.000
We can calculate the PGA at each point of interest,

00:44:30.000 --> 00:44:34.000
With common NGA West 2G ground motion models,

00:44:34.000 --> 00:44:37.000
Using the magnitude of the earthquake,

00:44:37.000 --> 00:44:39.000
And the distance to the fault.

00:44:39.000 --> 00:44:44.000
This fragile curve is for stationary trains, so using that PGA,

00:44:44.000 --> 00:44:49.000
would give us the derailment probability of a stationary train.

00:44:49.000 --> 00:44:52.000
At this point, I would like to mention that…

00:44:52.000 --> 00:44:55.000
To calculate the shaking intensity values,

00:44:55.000 --> 00:44:59.000
We used OpenSHA intensive measure calculation tool,

00:44:59.000 --> 00:45:07.000
Right? You serve 3 fall rupture geometries and magnitudes are already built in.

00:45:07.000 --> 00:45:13.000
Because of slowing the trains, the speed of trains with and without EW are obviously different.

00:45:13.000 --> 00:45:15.000
Without EW,

00:45:15.000 --> 00:45:21.000
the… we assume that the train speed is the nominal operating speed.

00:45:21.000 --> 00:45:27.000
And we also consider that strong shaking begins with the arrival of the S wave,

00:45:27.000 --> 00:45:33.000
at a given location, and except in the case of shear… super-shear rupture,

00:45:33.000 --> 00:45:38.000
The first S waves reaching a site originate from the hypocenter.

00:45:38.000 --> 00:45:47.000
Therefore, we calculate the arrival time by dividing the hypocentral distance by the…

00:45:47.000 --> 00:45:49.000
S-Wave, uh, velocity.

00:45:49.000 --> 00:45:51.000
And subtracting the time…

00:45:51.000 --> 00:45:55.000
required for earthquake detection and algorithm…

00:45:55.000 --> 00:45:58.000
Within processing from the S wave,

00:45:58.000 --> 00:46:01.000
Arrival time gives the time…

00:46:01.000 --> 00:46:03.000
Available for slowing the train.

00:46:03.000 --> 00:46:07.000
Multiplying this available time by the…

00:46:07.000 --> 00:46:11.000
Average deceleration gives the total speed

00:46:11.000 --> 00:46:15.000
reduction, and the train speed with EW…

00:46:15.000 --> 00:46:20.000
is obtained by subtracting the speed reduction

00:46:20.000 --> 00:46:25.000
from the operating train speed, so… so we end up with, uh, you know,

00:46:25.000 --> 00:46:31.000
the, uh, speed of a train when EW is in effect.

00:46:31.000 --> 00:46:37.000
But now, here, we direct to observe that the speed reduction depends on the hypocentral distance.

00:46:37.000 --> 00:46:39.000
So, for example, here,

00:46:39.000 --> 00:46:42.000
All the points of interest, uh, have shorter…

00:46:42.000 --> 00:46:48.000
Warning times for an earthquake originating near the middle of the fault compared to one.

00:46:48.000 --> 00:46:51.000
With hypocenter near the north,

00:46:51.000 --> 00:46:56.000
And because an earthquake can initiate from any location along the fault,

00:46:56.000 --> 00:47:01.000
You serve 3 rupture scenarios do not include hypocenter information.

00:47:01.000 --> 00:47:06.000
So, considering the importance of hypoential location,

00:47:06.000 --> 00:47:09.000
for EW in this study,

00:47:09.000 --> 00:47:13.000
We considered 5 different realizations.

00:47:13.000 --> 00:47:15.000
For each rupture with…

00:47:15.000 --> 00:47:19.000
Uniformly spaced hypocenters along the

00:47:19.000 --> 00:47:22.000
bolt rupture that you see with Starseer.

00:47:22.000 --> 00:47:28.000
Then after calculating the EW benefits for the entire BART system in each realization,

00:47:28.000 --> 00:47:33.000
We take the average of the five values to compute the EW benefit.

00:47:33.000 --> 00:47:35.000
for that rupture.

00:47:35.000 --> 00:47:39.000
As I mentioned, this fragile curve is for a stationary train.

00:47:39.000 --> 00:47:44.000
Therefore, we convert the speed of the train to equivalent ground shaking,

00:47:44.000 --> 00:47:46.000
Using equations from the literature,

00:47:46.000 --> 00:47:48.000
And add that to PGA.

00:47:48.000 --> 00:47:54.000
So here, uh, the constant alpha is, uh, for, uh, that conversion.

00:47:54.000 --> 00:47:58.000
And we enter the fragility curves with the, you know, resulting values.

00:47:58.000 --> 00:48:00.000
With and without EW, and find the

00:48:00.000 --> 00:48:03.000
funding properties that

00:48:03.000 --> 00:48:05.000
Thanksgiving, uh…

00:48:05.000 --> 00:48:11.000
The reduction promo…

00:48:11.000 --> 00:48:15.000
in the program.

00:48:15.000 --> 00:48:28.000
Because…

00:48:28.000 --> 00:48:32.000
Lace.

00:48:32.000 --> 00:48:34.000
the path.

00:48:34.000 --> 00:48:47.000
these people to come.

00:48:47.000 --> 00:48:52.000
Cross.

00:48:52.000 --> 00:49:19.000
approval to your…

00:49:19.000 --> 00:49:26.000
location for each calculations for all producing a 7.3 magnitude.

00:49:26.000 --> 00:49:28.000
these benefits for approximate Greece.

00:49:28.000 --> 00:49:31.000
Uh, using this path on the right.

00:49:31.000 --> 00:49:36.000
Reduction of probability of derailment also reduces the cost of repair and replacement of cars.

00:49:36.000 --> 00:49:40.000
And we calculate the corresponding values using

00:49:40.000 --> 00:49:45.000
The left path, and we add these two values to calculate the EW benefit.

00:49:45.000 --> 00:49:47.000
for a given point of interest.

00:49:47.000 --> 00:49:52.000
Then we repeat these calculations for all of the points of interest.

00:49:52.000 --> 00:49:56.000
And sum them to calculate the EW benefit across the system.

00:49:56.000 --> 00:50:02.000
And multiplying this value by the probability of occurrence of the scenario earthquake,

00:50:02.000 --> 00:50:06.000
Results in the EEW benefit for that scenario.

00:50:06.000 --> 00:50:08.000
And as I mentioned, we calculated the…

00:50:08.000 --> 00:50:15.000
these benefits for approximately $185,000 ruptures.

00:50:15.000 --> 00:50:20.000
Application of the methodology results in a variety of outcomes.

00:50:20.000 --> 00:50:25.000
First, we can obtain the spatial distribution of any quantity for each scenario.

00:50:25.000 --> 00:50:30.000
As an example, here we see the results for one of the rupture scenarios.

00:50:30.000 --> 00:50:36.000
128 kilometer rupture in the East Bay of San Francisco Bay Area.

00:50:36.000 --> 00:50:40.000
Producing a 7.3 magnitude earthquake,

00:50:40.000 --> 00:50:43.000
With a hypocenter near the San Pablo Bay,

00:50:43.000 --> 00:50:45.000
In this particular realization.

00:50:45.000 --> 00:50:51.000
In terms of results, as the color changes from yellow to red, the value increases.

00:50:51.000 --> 00:50:55.000
So, for this particular realization, we can observe that

00:50:55.000 --> 00:50:58.000
The derailment probability reduction is highest.

00:50:58.000 --> 00:51:01.000
Near the tracks in the South Bay,

00:51:01.000 --> 00:51:06.000
Mainly because these locations experience high-intensity shaking,

00:51:06.000 --> 00:51:09.000
And they are farthest from the hypocenter.

00:51:09.000 --> 00:51:13.000
Meaning that they receive sufficient warning time,

00:51:13.000 --> 00:51:16.000
to slow down or stop the trains.

00:51:16.000 --> 00:51:18.000
These points here and here…

00:51:18.000 --> 00:51:22.000
In San Francisco, and Antioch are also far from the…

00:51:22.000 --> 00:51:28.000
High percenter, but the shaking intensity at these locations are not high.

00:51:28.000 --> 00:51:31.000
So that… the train…

00:51:31.000 --> 00:51:40.000
derailment probability is already low, even without EW, and slowing down the trains do not provide additional benefits.

00:51:40.000 --> 00:51:45.000
Now, on the right, we see the results from all ruptures.

00:51:45.000 --> 00:51:51.000
Here, the x-axis shows the highest shaking intensity in the system due

00:51:51.000 --> 00:51:54.000
to an earthquake, and the y-axis

00:51:54.000 --> 00:51:59.000
is the total EW benefit over a span of 50 years.

00:51:59.000 --> 00:52:03.000
So, I will not focus on the particular values here.

00:52:03.000 --> 00:52:09.000
However, it is interesting to observe that there is a distinct shaking intensity

00:52:09.000 --> 00:52:14.000
Around MMI 8.3, below which the EEI

00:52:14.000 --> 00:52:17.000
W benefits are minor.

00:52:17.000 --> 00:52:21.000
And there is a sharp increase in the benefits above this value.

00:52:21.000 --> 00:52:26.000
This is actually a direct consequence of the utilized fragility curve

00:52:26.000 --> 00:52:30.000
Where the probability of derailment is almost zero.

00:52:30.000 --> 00:52:35.000
below, uh, MMR 8.3…

00:52:35.000 --> 00:52:44.000
And, uh, so further slowing of trains provides little additional safety and economic benefit to the left of this point.

00:52:44.000 --> 00:52:48.000
So this mainly shows the importance of the…

00:52:48.000 --> 00:52:50.000
Utilize fragility curve,

00:52:50.000 --> 00:52:53.000
As it becomes central to the calculations.

00:52:53.000 --> 00:52:57.000
So that takes me to my last slide.

00:52:57.000 --> 00:53:04.000
So, at this point, the methodology is fully algorithmic, and we have implemented all calculations in an in-house developed code.

00:53:04.000 --> 00:53:12.000
So what this means is that we can update results conveniently using any new information, such as new fragility curves,

00:53:12.000 --> 00:53:16.000
obtained from future train tests or detailed numerical simulations,

00:53:16.000 --> 00:53:19.000
or any updates to fault rupture?

00:53:19.000 --> 00:53:24.000
And ground shaking intensity in the U.S. national seismic Hazard Model, or…

00:53:24.000 --> 00:53:29.000
Ground tracking intends to obtain from physics-based simulations that explicitly capture

00:53:29.000 --> 00:53:31.000
Rapture directivity and other effects.

00:53:31.000 --> 00:53:36.000
So, related to the last item I would like to highlight, the peer-simulated ground motion database.

00:53:36.000 --> 00:53:40.000
that we have been developing together with Lawrence Berkeley National Lab.

00:53:40.000 --> 00:53:43.000
And this database currently…

00:53:43.000 --> 00:53:45.000
Flu's physics-based ground motions for

00:53:45.000 --> 00:53:48.000
multiple realizations of a certain magnitude.

00:53:48.000 --> 00:53:55.000
Hayward fall, rupture, and wool soon expand to include additional Bay Area and other regional earthquakes scenarios.

00:53:55.000 --> 00:54:00.000
This is the link for the website database, if anyone isn't interested in that.

00:54:00.000 --> 00:54:10.000
And with that, I would like to thank you, and I look forward to any questions or…

00:54:10.000 --> 00:54:17.000
Thank you, Celine, for that very interesting talk. And that brings us to our last speaker today.

00:54:17.000 --> 00:54:26.000
Who is Ryan Kirsting from. Bueller Engineering, and I will talk about functional recovery provisions to support community resilience.

00:54:26.000 --> 00:54:29.000
So, Ryan, take it away.

00:54:29.000 --> 00:54:35.000
Good afternoon. I'm Ryan Kirsting, I'm a practicing instructional engineer with Buehler Engineering in Sacramento.

00:54:35.000 --> 00:54:40.000
In my presentation today will cover the development of the NEHERP functional recovery provisions,

00:54:40.000 --> 00:54:45.000
That will be contained within the 2026 NEHERP recommended seismic provisions.

00:54:45.000 --> 00:54:51.000
These functional recovery provisions will be a way for our codes and standards to better support our community resilience goals,

00:54:51.000 --> 00:54:56.000
And this presentation covers the work of the NEHERP Function Recovery Task Committee.

00:54:56.000 --> 00:55:01.000
I like to acknowledge the contributions of my co-chair, Dr. Abley Lyle.

00:55:01.000 --> 00:55:09.000
And all of the members of our committee, as well as the corresponding members of the committee who have contributed to this work over the past

00:55:09.000 --> 00:55:13.000
almost 4 years now.

00:55:13.000 --> 00:55:24.000
This presentation will cover a number of aspects of the functional recovery provisions, including how do we get here, and why do we need it, so those contributions and motivations preceding the work of the task committee itself.

00:55:24.000 --> 00:55:31.000
Then we'll dive a little deeper into what the provisions cover, some of the key terms and the concepts contained within the framework of these provisions.

00:55:31.000 --> 00:55:41.000
And then we'll take a look at where we're going beyond the work of the task committee and how these provisions can support community resilience on a larger scale.

00:55:41.000 --> 00:55:50.000
As a quick disclaimer, this presentation contains a combination of information and opinions, both of me, but also the work of the Task Committee.

00:55:50.000 --> 00:55:57.000
And that the work of the task committee and the Provisions Update Committee is still ongoing, so what you see here is still technically draft.

00:55:57.000 --> 00:56:01.000
Although we do know a version of what you see here today will be published.

00:56:01.000 --> 00:56:03.000
At the end, toward the end of this year.

00:56:03.000 --> 00:56:05.000
in the final publication.

00:56:05.000 --> 00:56:11.000
For those who may be unfamiliar with the recommended seismic provisions, they are developed by the Provisions Update Committee.

00:56:11.000 --> 00:56:19.000
Under the guidance of the Building Seismic Safety Council of the National Institute of Building Sciences, on behalf of the NEHERP agencies.

00:56:19.000 --> 00:56:25.000
The function Recovery Task Committee serves under the Provisions Update Committee.

00:56:25.000 --> 00:56:35.000
And for those who also may not be familiar, the NEHERP provision serve as a resource for code and standard organizations, for jurisdictions at local, state, and federal level.

00:56:35.000 --> 00:56:40.000
And for project teams who may be interested in implementing the newest and best practices for seismic design,

00:56:40.000 --> 00:56:44.000
Even if they haven't been yet adopted in the full building codes.

00:56:44.000 --> 00:56:49.000
So the NEHERP provisions do serve as a source document for the development of ASCE7 provisions,

00:56:49.000 --> 00:56:53.000
ASC7 is a reference standard in the International Building Code.

00:56:53.000 --> 00:57:02.000
and the International Building Code, at least here in California, serves as a source document for the California Building Code.

00:57:02.000 --> 00:57:11.000
So how do we get here? We obviously probably all know that community resilience is been a big topic of late, and the public wants and expects resilient communities, even from

00:57:11.000 --> 00:57:17.000
our current codes. So this focus on community resilience at all levels of government has been focused on

00:57:17.000 --> 00:57:20.000
Reducing costs and damage and losses.

00:57:20.000 --> 00:57:28.000
as also protecting and providing higher levels of safety, and in our case, in the wake, or in the threat of a damaging earthquake.

00:57:28.000 --> 00:57:36.000
We've actually recognized for some time the need to develop a better link between our community resilience goals and the way we go about designing.

00:57:36.000 --> 00:57:38.000
And constructing our buildings.

00:57:38.000 --> 00:57:41.000
In Lifeline infrastructure systems,

00:57:41.000 --> 00:57:51.000
This link has been acknowledged as needing to be some sort of measurement of time to recovery of function, not just basic safety or minimum

00:57:51.000 --> 00:57:55.000
safety. So, functional recovery performance objectives can serve as this link.

00:57:55.000 --> 00:58:01.000
And I'll just offer that we're going to use functional recovery as a measurement at a building or lifeline infrastructure scale.

00:58:01.000 --> 00:58:07.000
In support of resilience that can be measured at a community scale.

00:58:07.000 --> 00:58:11.000
A little bit more about why we need to think about functional recovery performance.

00:58:11.000 --> 00:58:22.000
Partly our… we just need to acknowledge that our current codes and standards do not explicitly protect against economic loss, nor target performance in terms of return of function for most of our buildings.

00:58:22.000 --> 00:58:27.000
Recent studies by FEMA and ATC over the last 3 to 5 years

00:58:27.000 --> 00:58:30.000
Maybe even longer ago, now that I think about it.

00:58:30.000 --> 00:58:44.000
would expect that 20% or more of modern code-conforming buildings may be unfit for occupancy or function for months or longer following a major earthquake. So, if we're trying to protect our individuals and our communities against

00:58:44.000 --> 00:58:49.000
Such losses and long recovery times, we need to acknowledge that we should be changing the way we…

00:58:49.000 --> 00:58:54.000
approach our codes and standards.

00:58:54.000 --> 00:59:03.000
So a lot of that information and those desires and those understandings led to some of the language that was included in 2018 NEHERP reauthorization.

00:59:03.000 --> 00:59:08.000
In which Congress charged the NEHERP agencies to consider community resilience goals when…

00:59:08.000 --> 00:59:10.000
They were taking activities.

00:59:10.000 --> 00:59:15.000
regulating to development of codes and standards for seismic design.

00:59:15.000 --> 00:59:24.000
The bill also specifically charged FEMA and NIST to form a committee of experts to provide recommended options for how we can improve the built environment

00:59:24.000 --> 00:59:29.000
Relative to the post-earthquake reoccupancy and functional recovery time.

00:59:29.000 --> 00:59:35.000
So this resulted in a 2021 FEMA NIST report to Congress, as we like to shorthand it. I was involved in it

00:59:35.000 --> 00:59:39.000
development of that report, and now I've also been involved in the efforts.

00:59:39.000 --> 00:59:46.000
that have been, uh, charged by NIBs and BSSC to the PUC to consider functional recovery in the

00:59:46.000 --> 00:59:50.000
development of the 2026 NEHERP provisions.

00:59:50.000 --> 00:59:57.000
A quick summary of the FEMA NIST report. It came up with 7 primary recommendations related to community resilience and how do we improve

00:59:57.000 --> 01:00:00.000
post-earthquake function recovery time,

01:00:00.000 --> 01:00:10.000
Recognizing that these seven recommendations do not just apply to any particular aspect of community resilience, we really need to work on all of these almost simultaneously, or at least in

01:00:10.000 --> 01:00:18.000
Parallel. The work of the functional recovery task committee in general is focused primarily on Recommendations 1 or 2.

01:00:18.000 --> 01:00:26.000
1 and 2, developing the framework for these recovery-based objectives, and then also considering how to design new buildings and other structures.

01:00:26.000 --> 01:00:30.000
to meet those recovery-based objectives.

01:00:30.000 --> 01:00:42.000
So, in a quick summary, there was a lot of other work that I haven't been able to get into, even on this page, it's not listing all of it, but I just want to acknowledge that there has been a lot of other relevant work and resources to get us to where we are today.

01:00:42.000 --> 01:00:47.000
that have enabled us to get to the point of developing functional recovery provisions.

01:00:47.000 --> 01:00:50.000
Uh, as approaching the way we design

01:00:50.000 --> 01:00:53.000
new buildings.

01:00:53.000 --> 01:00:58.000
So, a little bit more now about function recovery, and how do we define it, and what is it.

01:00:58.000 --> 01:01:13.000
Function recovery, as a definition, is given as a post-event performance state in which a building, or maybe a lifeline infrastructure system, is maintained or restored to support the basic intended functions associated with its pre-event use.

01:01:13.000 --> 01:01:19.000
The important part to note here is that basic intended functions are less than full functionality.

01:01:19.000 --> 01:01:24.000
But more than what would be considered minimum for reoccupancy. So, in general…

01:01:24.000 --> 01:01:31.000
a building, if you were applying this to a building, would be ready to support most of its use, but not all of its use.

01:01:31.000 --> 01:01:34.000
In addition to reoccupancy.

01:01:34.000 --> 01:01:39.000
If that's the performance state, you could consider a functional recovery performance objective,

01:01:39.000 --> 01:01:41.000
being… achieving that.

01:01:41.000 --> 01:01:46.000
state within an acceptable time following a specified hazard event.

01:01:46.000 --> 01:01:51.000
Where that acceptable time might differ for various building uses or occupancies or services.

01:01:51.000 --> 01:01:56.000
So, functional recovery in itself involves recovery of functions…

01:01:56.000 --> 01:02:00.000
When our communities need the services provided by that building.

01:02:00.000 --> 01:02:07.000
Function recovery does not meet immediate recovery for all uses, occupancies, or services.

01:02:07.000 --> 01:02:19.000
So just to be crystal clear, we're going to make sure we don't equate these terms… functional recovery is not going to be equivalent to immediate occupancy, nor immediate recovery. Function recovery is not equivalent to full recovery.

01:02:19.000 --> 01:02:22.000
For those in the weeds of ASCE 7 with me.

01:02:22.000 --> 01:02:30.000
There are some terms about functionality already in ASC7, and functional recovery is not the same as those. There's some subtle nuances there that we can talk about if

01:02:30.000 --> 01:02:32.000
If we need to.

01:02:32.000 --> 01:02:41.000
Uh, and then functional recovery, again, is going to be measured at a building or asset level, whereas resilience is going to be measured at a community level. So function recovery supports

01:02:41.000 --> 01:02:48.000
community resilience, but it's not quite the same as community resilience.

01:02:48.000 --> 01:02:51.000
The Function Recovery Task Committee in general…

01:02:51.000 --> 01:03:00.000
was motivated by a number of goals. The primary ones were simply to reduce our vulnerability, to minimize losses, and to improve recovery time.

01:03:00.000 --> 01:03:02.000
If you start to wade a little deeper into those,

01:03:02.000 --> 01:03:11.000
What does that mean? Well, we were thinking of reducing total cost of ownership of these facilities, total cost being not just initial construction costs, but long-term

01:03:11.000 --> 01:03:15.000
operating costs as well as maintenance and repair costs following an event.

01:03:15.000 --> 01:03:21.000
And then also improving access to essential services. With essential services now being something

01:03:21.000 --> 01:03:25.000
understood to be something more than just emergency services, thinking of…

01:03:25.000 --> 01:03:28.000
daily and weekly necessities as well.

01:03:28.000 --> 01:03:33.000
And knowing that one of the key aspects of a resilient community is also to avoid significant

01:03:33.000 --> 01:03:40.000
population displacement. So, understanding we also need to have reasonable recovery time for places like

01:03:40.000 --> 01:03:42.000
schools, uh, workplaces,

01:03:42.000 --> 01:03:49.000
And also make sure that we consider the potential for cascading effects of the performance of multifamily housing.

01:03:49.000 --> 01:03:54.000
Housing, obviously, being a very critical aspect of a resilient community.

01:03:54.000 --> 01:04:07.000
Another kind of longer-term or broader goal was recognizing that the… what we were doing was really developing first-generation code language, but as we were doing so, we wanted to make sure that what we were doing was not just achieving some short-term goal.

01:04:07.000 --> 01:04:16.000
Um, we wanted to be setting up the stage, building something that we could use to ultimately support broader community resilience.

01:04:16.000 --> 01:04:19.000
through future seismic building code provisions.

01:04:19.000 --> 01:04:24.000
Recognizing that what we're doing with each building we design and construct

01:04:24.000 --> 01:04:28.000
It really does have impacts across the entire community. So,

01:04:28.000 --> 01:04:35.000
Uh, we do recognize that we are working on this first generation, and we will continue to develop and refine these

01:04:35.000 --> 01:04:38.000
to further support community resilience goals.

01:04:38.000 --> 01:04:41.000
In future editions.

01:04:41.000 --> 01:04:43.000
Couple things more now about the appendix itself.

01:04:43.000 --> 01:04:50.000
First of all, this is a voluntary appendix. It does contain design criteria and other provisions.

01:04:50.000 --> 01:04:59.000
that reflect the latest research and technical studies, uh, that reflect and consolidate current best practices already being used for the designing and construction of

01:04:59.000 --> 01:05:04.000
buildings. We wanted to make sure to consolidate those for consistency.

01:05:04.000 --> 01:05:12.000
And this voluntary appendix is something that can be considered for implementation. You don't have to wait for it to get into ASC7 or the building code.

01:05:12.000 --> 01:05:17.000
Most jurisdictions have the ability to consider implementing enhanced

01:05:17.000 --> 01:05:19.000
requirements for their communities if needed.

01:05:19.000 --> 01:05:24.000
And these provisions are written in a way that a community could consider adopting them.

01:05:24.000 --> 01:05:31.000
for implementation. And even if a community decided not to use them, you could still implement them at an individual…

01:05:31.000 --> 01:05:34.000
facility level, or an individual owner level.

01:05:34.000 --> 01:05:37.000
So, some of the key aspects of the framework.

01:05:37.000 --> 01:05:43.000
I'm not going to get into a lot of design criteria today, but I want to at least describe the framework and how…

01:05:43.000 --> 01:05:46.000
Communities can engage with these provisions. So,

01:05:46.000 --> 01:05:57.000
These provisions use a new concept called a functional recovery category to differentiate target recovery times and identify related design criteria and provisions to achieve those target times.

01:05:57.000 --> 01:06:04.000
Again, reminding you that target times can be selected to meet your overall community resilience goal for that particular facility.

01:06:04.000 --> 01:06:13.000
And understanding to target the recovery time for when those functions and services are needed in the overall community recovery spectrum.

01:06:13.000 --> 01:06:16.000
Function recovery categories are different than risk categories. There may be some…

01:06:16.000 --> 01:06:20.000
Risk category 2 facilities that actually need quick recovery.

01:06:20.000 --> 01:06:27.000
Uh, and they might be lumped into a recovery… a functional recovery category with other risk category 4 facilities.

01:06:27.000 --> 01:06:35.000
Also, just understand that when we are using these or developing this design criteria, we were considering a risk-targeted performance approach,

01:06:35.000 --> 01:06:40.000
with a certain risk and reliability target of a 10% probability of exceeding

01:06:40.000 --> 01:06:43.000
These recovery times in a 50-year lifespan of a building.

01:06:43.000 --> 01:06:46.000
And that just for those who are interested,

01:06:46.000 --> 01:06:50.000
The ground motion that allows us to achieve that target

01:06:50.000 --> 01:07:00.000
would be something a little bit smaller than what we use currently for safety-based design. So, a functional recovery earthquake, if you were to think of it that way, is a slightly more frequent, although slightly smaller, earthquake.

01:07:00.000 --> 01:07:06.000
than compared to what we're using for safety-based design. But it achieves this 10% probability of

01:07:06.000 --> 01:07:09.000
exceeding that, uh, recovery time.

01:07:09.000 --> 01:07:14.000
The recovery categories themselves are arranged more in a letter grade style,

01:07:14.000 --> 01:07:16.000
So, A being the best, uh, and then…

01:07:16.000 --> 01:07:21.000
transitioning lower with longer recovery times as you work through B, C, and D.

01:07:21.000 --> 01:07:25.000
We do have a Category E for which you don't have to consider it.

01:07:25.000 --> 01:07:32.000
And again, the degree of urgency for the types of facilities would be the most immediate would be an A.

01:07:32.000 --> 01:07:36.000
And then urgent would be in B, and then some of the…

01:07:36.000 --> 01:07:45.000
you know, short and medium-term recovery would be in C, and then the longer return to normal would be in category D.

01:07:45.000 --> 01:07:51.000
So we have that general framework in the provisions, and then we offer a commentary on where certain…

01:07:51.000 --> 01:07:56.000
types of facilities might be assigned, and we leave the actual assignment of this to

01:07:56.000 --> 01:07:59.000
a community or an owner to decide.

01:07:59.000 --> 01:08:02.000
But Category A, again, if you're thinking about…

01:08:02.000 --> 01:08:09.000
recovery within a few hours would be those emergency services, but we also recommend thinking about housing for

01:08:09.000 --> 01:08:12.000
Um, vulnerable populations.

01:08:12.000 --> 01:08:16.000
Category B has a recovery time of a few days.

01:08:16.000 --> 01:08:20.000
And you'll see here, urgent services, as well as food and water services.

01:08:20.000 --> 01:08:25.000
and housing would land here. And I guess this is a good time to interject that…

01:08:25.000 --> 01:08:31.000
Uh, the group that helped develop these recommendations, uh, represented a broad range of expertise.

01:08:31.000 --> 01:08:36.000
from, uh, economic advice to community resilience advocates to,

01:08:36.000 --> 01:08:46.000
Um, those who know social scientists much better than I do as a structural engineer, but we relied on that expertise and knowledge to inform us of these recommendations.

01:08:46.000 --> 01:08:49.000
Category C of that targeted within a couple of months.

01:08:49.000 --> 01:08:52.000
Here's where you'll notice a little bit more of the community.

01:08:52.000 --> 01:08:56.000
facilities, as well as education facilities.

01:08:56.000 --> 01:08:58.000
And then Category D…

01:08:58.000 --> 01:09:04.000
would be those that are longer term.

01:09:04.000 --> 01:09:07.000
And obviously, I don't have the time to get into all the different examples here, but…

01:09:07.000 --> 01:09:11.000
But again, the idea here is that this is a voluntary appendix.

01:09:11.000 --> 01:09:19.000
With recommendations that a community could choose, and they could choose to modify. And we want to make sure we gave them some recommendations, but also gave them the option

01:09:19.000 --> 01:09:22.000
for that flexibility.

01:09:22.000 --> 01:09:28.000
So, sort of in closing here, where are we going from here? Again, these provisions will be published at the end of this year.

01:09:28.000 --> 01:09:34.000
They'll be available for organizations or jurisdictions who might be interested in looking for enhanced

01:09:34.000 --> 01:09:37.000
performance, and beyond just the basic code safety.

01:09:37.000 --> 01:09:41.000
And so they could be implementing functional recovery provisions included in the NEHERP

01:09:41.000 --> 01:09:47.000
recommended provisions. They could choose to do that at a community scale, they could choose to do it

01:09:47.000 --> 01:09:50.000
For certain types of buildings in their community, they could choose to do it.

01:09:50.000 --> 01:09:57.000
As an owner, for some or more of their own facilities.

01:09:57.000 --> 01:09:59.000
And then just a little personal look ahead.

01:09:59.000 --> 01:10:06.000
We know these are going to be published this year, we know this first edition makes really important advances. We know we'll continue to build on this.

01:10:06.000 --> 01:10:11.000
But I don't think these concepts are too foreign, and so I think, really, the idea here is that…

01:10:11.000 --> 01:10:14.000
It's gonna look awfully similar, it just allows us that…

01:10:14.000 --> 01:10:23.000
opportunity to start considering functional recovery performance and the use of category to achieve the different recovery times should make it easy to implement.

01:10:23.000 --> 01:10:25.000
Our goal is that we've made it

01:10:25.000 --> 01:10:33.000
Something easy enough for leading jurisdictions, organizations, firms, engineers to begin implementing or at least considering.

01:10:33.000 --> 01:10:42.000
So, in areas of high seismic risk and vulnerability, or for projects where there's an intentional desire for higher performance, I think these are… these provisions would be ready to be used.

01:10:42.000 --> 01:10:47.000
If nothing else, we hope the presence and availability of these provisions

01:10:47.000 --> 01:10:52.000
Provide additional context and conversations so that decisions can be informed.

01:10:52.000 --> 01:10:59.000
by better understanding of the recovery time performance one might want or need.

01:10:59.000 --> 01:11:01.000
In the longer term,

01:11:01.000 --> 01:11:12.000
Right now, this is an additional check you would do in addition to a life safety check, but we believe that we can transition this to be the primary check as the standard and singular design requirement in a longer term.

01:11:12.000 --> 01:11:20.000
And we also think longer-term that what we're doing for new buildings can inform what we're doing eventually for a more existing buildings as well.

01:11:20.000 --> 01:11:23.000
We know that these functional recovery performance…

01:11:23.000 --> 01:11:26.000
Objectives and these new provisions will ultimately improve the

01:11:26.000 --> 01:11:31.000
resilience for our communities, we just need to make sure we get the word out that there's starting to be available.

01:11:31.000 --> 01:11:33.000
for people to consider.

01:11:33.000 --> 01:11:39.000
So with that, I think we're out of time. There are some additional resources here, and I'm more than happy to answer some questions.

01:11:39.000 --> 01:11:44.000
Thank you very much.

01:11:44.000 --> 01:11:49.000
Thank you, Ryan, for… Walking us through the NIH provisions.

01:11:49.000 --> 01:11:58.000
Um, so, it's time for the Q&A. And you can raise your hand if you want to ask your question verbally.

01:11:58.000 --> 01:12:02.000
In the meantime, I'll go through a couple of questions for John that we're not answered.

01:12:02.000 --> 01:12:07.000
in the chat, or at least I think they were not answering the chat.

01:12:07.000 --> 01:12:12.000
So, uh, the first is from David. So, John.

01:12:12.000 --> 01:12:17.000
is the social modeling and decision-making based only on a general closed-form model?

01:12:17.000 --> 01:12:24.000
Or are you doing something like agent-based modeling? In my experience, there's a wide variety in actual decisions.

01:12:24.000 --> 01:12:28.000
Even for similarly situated people, households.

01:12:28.000 --> 01:12:57.000
Yeah, so the, um, so the great question. So, sorry I missed that, but, um, so the… The model itself, the CGE model, of course, is a, uh, uses a social accounting matrix. It has, you know, hundreds, sometimes if it's a big community, can have even more interacting equations, and these are… and it basically, uh, basically, uh, you know, it's, it's done by our team of economists, not by the engineers, obviously.

01:12:57.000 --> 01:13:10.000
Um, for the population dislocation, things like that. That one's actually a logistic regression based on, uh, about 20 to 25 years' worth of field studies, including our Lumberton study.

01:13:10.000 --> 01:13:30.000
And, uh, and primarily, a lot of those are post-hurricane, and I think only one is… so we've sort of melded some of the, uh, you know, so it's based on damage, it's based on race ethnicity, it's based on income, as well as a few other things, and so… so it's not, uh, not just base.

01:13:30.000 --> 01:13:40.000
solely on damage. So that's the dislocation parameter. But for the CGE, it also takes that into account, so it's, uh, so it's taking into account.

01:13:40.000 --> 01:14:07.000
Um, uh, you know, a lot of things, but they, they use things like elasticities and things, so there's really some, for the economists, there's some finesse to it, but it's, uh, the CG is probably, you know, arguably the economists at least inform us, it's the, probably the most powerful approach to modeling economic state following a shock, as they put it, in this case, an earthquake.

01:14:07.000 --> 01:14:11.000
Okay. So then, Sarah Minson also has a question.

01:14:11.000 --> 01:14:18.000
Uh, so, as someone who is familiar with the building stock in the northern coast, is your building inventory in Rio del.

01:14:18.000 --> 01:14:23.000
just official permitted buildings, or does it include the other things that people build?

01:14:23.000 --> 01:14:28.000
And, um… Use for homes and business.

01:14:28.000 --> 01:14:45.000
Yeah, and I made an attempt at that one, but probably not a great one. So for that, um, the, uh, we started with the, we started with the National Building Inventory, but then cross-check, so anything that… anything that was inspected, uh, and… and tagged by.

01:14:45.000 --> 01:15:04.000
Uh, by the inspections, um, post-earthquake was… Uh, was modeled. However, when we got to the other communities, there had been no inspection because there hadn't been an earthquake, or at least the effects of an earthquake, so… so those just use the NBI, but we did cross-check that, because there's people that.

01:15:04.000 --> 01:15:26.000
use that. No, um, sometimes there's a, uh, sort of a random ML error that appears, or something, so… Um, so we made sure with, uh, some… uh, rough visual checks, but for the most part, pretty accurate. There weren't that, you know, the small community, so it wasn't too hard to check.

01:15:26.000 --> 01:15:36.000
Okay, uh… I encourage people in the… who are in the call to raise their hand and ask any questions to any of our.

01:15:36.000 --> 01:15:43.000
of our speakers, if there… if there isn't any… I can… I can ask one for… another one for you, John.

01:15:43.000 --> 01:15:58.000
So, I'm not familiar with the InCore modeling. And, uh, you showed… you showed, uh, result for some scenarios, some scenario earthquakes. I was… I was wondering if there were… if there were capabilities to.

01:15:58.000 --> 01:16:03.000
kind of get risk results, maybe couple that with the National Seismic Hazard Model.

01:16:03.000 --> 01:16:21.000
Yeah, great, great question, Ellen. So, um, so… so the, the, um, for that, um, right now, uh, when it was developed, it was developed as a, uh, sort of started as a research tool, uh, in collaboration with NIST, and then, uh, and then, uh, evolved, uh, during the last.

01:16:21.000 --> 01:16:43.000
5 years evolved into a, uh, planning tool, but a technical planning tool. And, uh, and so the… For communication, it was done with scenarios. However, it's, you know, the actual, uh, the actual code, everything is Jupyter notebooks, it's all Python-based, and so… so we do have, you know, sometimes students or postdocs that will.

01:16:43.000 --> 01:17:04.000
Uh, we'll go ahead and look at, um… look at, uh, say, a number of scenarios to then, um, to then link it with risk, and I would say, uh, I can't remember who did that, it might have been Jack Baker, there was an approach, I think I was… I think I saw Dustin on here, we were talking about that the other day, but that would probably be the way to do it, to go ahead and.

01:17:04.000 --> 01:17:10.000
Uh, try and reproduce a hazard curve with a number of scenarios, uh, and.

01:17:10.000 --> 01:17:14.000
Uh, and then you would get your risk, and so… so that would really just be putting another.

01:17:14.000 --> 01:17:19.000
loop on the outside as you run it.

01:17:19.000 --> 01:17:21.000
Okay. Sarah, you have your hand up.

01:17:21.000 --> 01:17:29.000
Hey! So, unfortunately, this is probably another, um… Question for John, putting John on the spot, and I was wondering, as you looked at your comparison between.

01:17:29.000 --> 01:17:33.000
The Odell and the other two communities is also kind of a rural.

01:17:33.000 --> 01:17:40.000
Or then divide between those in addition to the other economic and demographic issues, and I was wondering if you.

01:17:40.000 --> 01:17:50.000
had thought about that at all, because Alessandra was trying to educate me a bit on some of these differences, which is that, you know, on the one hand, for example, if you're in a rural area, you may be less likely to get, uh.

01:17:50.000 --> 01:17:57.000
disaster declaration that would assist with recovery. On the other hand, if you're in a rural area, you're probably more used to having to fix things on your own, right?

01:17:57.000 --> 01:18:06.000
If the county's not going to come out to clear your road when the tree falls down for regular… you know, storms, and you're used to getting out the chain song, clearing the road when it comes down from an earthquake.

01:18:06.000 --> 01:18:19.000
Yeah, Sarah, that's a great, uh, great comment. We did debate that, uh, that issue, and what really sort of sealed, you know, selecting those communities, we went through quite a few was.

01:18:19.000 --> 01:18:31.000
Uh, was really the number of out-commuters, so realizing that while they're rural, they're still out commuting and going to work, at least 90-some percent. And so the economist felt that that.

01:18:31.000 --> 01:18:40.000
That would… that would really tell the tale from an economic recovery. Um, however, what you're saying is, you know, absolutely true. I mean, I'm thinking of.

01:18:40.000 --> 01:18:52.000
you know, when we try and model the recovery of communities, and we've modeled the building recovery, and we're always off right at first, like, after tornadoes and any event, really, any hazard.

01:18:52.000 --> 01:19:06.000
the first 6 months, our estimates are usually… our predictions are off, and that's because people fix things themselves. Uh, and so… and we can't model that. We can't model if somebody's gonna pick up a hammer and go out on a Saturday.

01:19:06.000 --> 01:19:22.000
And so, um, so that's a great point. I'd say, you know, I said apples to apples. I guess maybe it's like a, I don't know, a Macintosh to a, you know, Red Delicious, I'm not sure, but it's close.

01:19:22.000 --> 01:19:31.000
I have two questions pending right now. One is Andy Michael in the chat, who's raised their hand, and then after that, I'll try and insert, um.

01:19:31.000 --> 01:19:37.000
Rajivani's more technical, like, workflow. data integration question after that.

01:19:37.000 --> 01:19:41.000
It's good, thanks. I want to actually follow up on Sarah's point.

01:19:41.000 --> 01:19:46.000
So, in 1992, when there were the Cape Mendocino earthquakes, I actually wound up in Ferndale, which is about

01:19:46.000 --> 01:19:49.000
15 miles from Rio Del,

01:19:49.000 --> 01:19:53.000
I actually checked during the talk, uh, population is smaller,

01:19:53.000 --> 01:19:59.000
income is about medium income right now is around $60,000, so it's not… it's more similar to Rio Del than it is to anything else.

01:19:59.000 --> 01:20:00.000
Hmm.

01:20:00.000 --> 01:20:07.000
I remember seeing the Red Cross up there, um, we got to look at the entrance, and Rich Eisner, who was then the head of the Earthquake program, Cal OES, and…

01:20:07.000 --> 01:20:10.000
Yeah, the community was very resilient.

01:20:10.000 --> 01:20:14.000
People had already been to the lumberyard, people were already jacking homes back up,

01:20:14.000 --> 01:20:17.000
to, you know, these are Victorians back onto their foundations, and…

01:20:17.000 --> 01:20:21.000
It was sort of almost a shrug our shoulders. I mean, we were getting lots of…

01:20:21.000 --> 01:20:23.000
not science, but from the emergency…

01:20:23.000 --> 01:20:25.000
From a response point of view, people were sort of…

01:20:25.000 --> 01:20:27.000
stunned by, um…

01:20:27.000 --> 01:20:34.000
Yeah, they weren't as needed as they expected. Other factor was a lot of people owned RVs or had families in town, so people were just…

01:20:34.000 --> 01:20:39.000
Either living in their… they were already… they were already setting up to live in their driveways, or…

01:20:39.000 --> 01:20:44.000
In an neighbor's house. So, it was a very different situation.

01:20:44.000 --> 01:20:57.000
Yeah, and thanks, Sandy, and that's, um, you know, we see the same thing, you know, I think you guys know, all… a lot of you know Judy Matroni Reiser, you know, did the study down in New Zealand and saw the same thing with.

01:20:57.000 --> 01:21:22.000
uh, you know, staff at hospitals after Christchurch and, uh, and so they, they see, and really, the, you know, the big conclusion is, you know, that the physical environment's easy to model, it's the humans that are hard, and we run into that a lot, and of course, you know, the standard deviations or the variance on a lot of those models are quite large, and so we're forcing sort of the mean.

01:21:22.000 --> 01:21:46.000
in for those when we're using variation in some of the engineering models. So, yeah, your points, you know, we definitely see that. We see that a lot after tornadoes as well, and a little less so with floods. I think that one is really difficult for people, simply because of the mess, but yeah, so thanks for the comment, I agree.

01:21:46.000 --> 01:21:59.000
All right, um, so… Rajpijani's question relates to my dog barking, clearly. How can GIS and remote sensing strengthen seismic risk.

01:21:59.000 --> 01:22:06.000
mapping workflows, and I would just broaden that to say all of the talks today involved remarkable levels.

01:22:06.000 --> 01:22:15.000
of data integration across. various modeling platforms and phenomenon and so on. So, anyone can answer.

01:22:15.000 --> 01:22:22.000
Uh, in either the specific or the broader of the challenges of bringing these data.

01:22:22.000 --> 01:22:30.000
together in a cohesively usable way.

01:22:30.000 --> 01:22:31.000
Go, David. Yes.

01:22:31.000 --> 01:22:38.000
Shall I go ahead? So, um, I'm not sure this is exactly what he's asking, but in my work with cities.

01:22:38.000 --> 01:22:42.000
I want all the decision-making to be based on.

01:22:42.000 --> 01:22:47.000
the data, and the data is a building inventory. So the question has been.

01:22:47.000 --> 01:22:54.000
Can we use some kind of an AI or, you know, Google imagery that exists to actually create an inventory?

01:22:54.000 --> 01:22:58.000
And it's a great idea. I have not yet seen it work.

01:22:58.000 --> 01:23:03.000
And much of that is because you can't tell the structural information.

01:23:03.000 --> 01:23:05.000
from what you can see from the street view.

01:23:05.000 --> 01:23:10.000
And you can't go around the back, so… Someday that may happen.

01:23:10.000 --> 01:23:14.000
But… and I'm very supportive of efforts to try to make it work.

01:23:14.000 --> 01:23:24.000
But it hasn't really been useful yet. It might work at a very broad scale. John mentioned that you're using, um.

01:23:24.000 --> 01:23:25.000
Uh, yeah, national building, or national structure inventory, but yeah.

01:23:25.000 --> 01:23:28.000
what's the National Building Inventory, what's it called?

01:23:28.000 --> 01:23:34.000
Yeah, and if you have… if you're trying to reach broad conclusions about a whole city, maybe you can smear that data, or.

01:23:34.000 --> 01:23:37.000
Or think of it as, uh, you know, good enough in the aggregate.

01:23:37.000 --> 01:23:42.000
But when you're making decisions about a couple of hundred buildings within a community.

01:23:42.000 --> 01:23:45.000
Uh, we need something a little more specific.

01:23:45.000 --> 01:23:57.000
Yeah, and maybe I'll make the point, David, on top of that, that, you know, like, one of the ways we do it, I mean, yeah, if that information was readily available, it would really speed things up. I mean.

01:23:57.000 --> 01:24:03.000
When we, uh, we first studied Joplin about a decade ago, um, after a tornado… after the big tornado there.

01:24:03.000 --> 01:24:30.000
We, uh, we actually assigned the 30,000 buildings, we assigned the archetypes manually, so a student went through and did those individually. There was no building inventory, there was no… You know, and much of what they had were still paper records, and so we said, well, we can… we can do that, we can drive the streets, and, uh, for tornadoes, it's, you know, less important knowing what the structural system is, a little less, but, uh, so, uh, David.

01:24:30.000 --> 01:24:36.000
I will say that in the inventories, I mean, I'm doing an inventory for a Bay Area city right now.

01:24:36.000 --> 01:24:40.000
I do it 100% from my desk. Because I had that imagery, but it's still… it's a combination of the imagery and the human eyeballs.

01:24:40.000 --> 01:24:44.000
Yeah.

01:24:44.000 --> 01:24:49.000
So it's not fully automated yet, and I don't think it would be that much faster to actually try to do it.

01:24:49.000 --> 01:24:52.000
by AI, if you still have a lot of checking to do.

01:24:52.000 --> 01:24:57.000
But the combination is absolutely pivotal and something we did not have.

01:24:57.000 --> 01:24:59.000
when we were doing sidewalk surveys 20 years ago.

01:24:59.000 --> 01:25:00.000
Because you can't sit on the back.

01:25:00.000 --> 01:25:25.000
Well, so in the… Yeah, yeah, and the National Structure Inventory would, you know, just sort of following up, the, uh, the U.S. Army Corps, they… what they did was they basically went through, they gathered from all of the counties, all of the, you know, the entities, they gathered everything they had, everybody that was willing to give it, and then… and then the rest, uh, there were little gaps, especially in the rural areas, and so that was done.

01:25:25.000 --> 01:25:39.000
with an ML approach, like, sort of an unspecified, as far as I know, I don't know the details of anybody from USGS is on, but… Um, and so we do see some errors, particularly in the rural areas, and those can get a little squirrely.

01:25:39.000 --> 01:25:41.000
I'd say even in the urban areas, too, I mean… The assessor's data is great for what assessors need to do.

01:25:41.000 --> 01:25:46.000
Yeah, oh yeah.

01:25:46.000 --> 01:25:49.000
It's not great for what engineers need to know.

01:25:49.000 --> 01:25:55.000
Right, right, I think we've come across maybe, like, a 60-story school, and we're like, I don't think so.

01:25:55.000 --> 01:26:03.000
So there are a lot of USGS folks in the talk, but I think we're gonna… gonna pivot to ask a question to one of our other.

01:26:03.000 --> 01:26:21.000
speakers. So, um… I don't see anything in the chat, so I'll ask a question, so… Selim, uh… I was a very, very interesting study, and something that I… I didn't know if you… I don't know if you said this or not, or maybe I missed it.

01:26:21.000 --> 01:26:28.000
But, uh, it seems like that fragility curve is very essential in all the computations, and I didn't… I didn't get.

01:26:28.000 --> 01:26:32.000
where that fragility came from. You mentioned the literature, but is that based on.

01:26:32.000 --> 01:26:36.000
Is that based on testing, or data? How is that fragility function?

01:26:36.000 --> 01:26:44.000
Yeah, actually, that's a great question. I… I can send you the link for that particular study, so as I mentioned, it was

01:26:44.000 --> 01:26:46.000
you know, sort of challenging to find.

01:26:46.000 --> 01:26:49.000
Those type of fragilities, and that particular one is…

01:26:49.000 --> 01:26:52.000
you know, a study from China,

01:26:52.000 --> 01:27:00.000
And they have modeled… they don't explicitly, you know, mention a specific train.

01:27:00.000 --> 01:27:02.000
But then they give all the dimensions and all that.

01:27:02.000 --> 01:27:08.000
And the, you know, the fragility curve uses a certain damage state.

01:27:08.000 --> 01:27:13.000
Uh, which is, uh, you know, offlifting a certain amount of the…

01:27:13.000 --> 01:27:17.000
wheel from the track is what they call as, you know, the derailment.

01:27:17.000 --> 01:27:20.000
And that is the damage state.

01:27:20.000 --> 01:27:25.000
And it's all obtained from, you know, simulations, not any testing.

01:27:25.000 --> 01:27:29.000
But, uh, actually, we in the lab…

01:27:29.000 --> 01:27:30.000
I want to do that.

01:27:30.000 --> 01:27:31.000
Okay.

01:27:31.000 --> 01:27:34.000
may have the opportunity to test the real train.

01:27:34.000 --> 01:27:40.000
So, in the peer labs, so that will be exciting, too.

01:27:40.000 --> 01:27:41.000
So that will be coming.

01:27:41.000 --> 01:27:42.000
Cool.

01:27:42.000 --> 01:27:46.000
Can the, uh, can the panel ask each other questions as well, or, uh…

01:27:46.000 --> 01:27:48.000
Do you have a question for Salim?

01:27:48.000 --> 01:27:49.000
Yeah, I…

01:27:49.000 --> 01:27:59.000
That's what we want to do make sure to get some questions in for Saleem and Ryan, too, so…

01:27:59.000 --> 01:28:00.000
Yep, go for it.

01:28:00.000 --> 01:28:12.000
Yeah, I guess I had a question for Saleem on the train, if that's okay, but… Um, so… so… so it seems like you… following up on Alan's on the fragility, Salim, it seems like, um, you mentioned that it was also a function of train velocity, so is it.

01:28:12.000 --> 01:28:13.000
So, yeah, that's what I needed to be, sort of, you know, creative.

01:28:13.000 --> 01:28:17.000
So is there… so is for each velocity, is there a fragility, or…

01:28:17.000 --> 01:28:22.000
Because that one, I tried to, you know, explicitly mention that, but maybe it wasn't so clear.

01:28:22.000 --> 01:28:28.000
That fragility that I showed was being used for a stationary train, meaning, you know, there's no speed.

01:28:28.000 --> 01:28:29.000
But ideally, you know, to consider the effect of

01:28:29.000 --> 01:28:32.000
Oh, okay.

01:28:32.000 --> 01:28:34.000
speed, you would want to have…

01:28:34.000 --> 01:28:40.000
Fragility curves for different train speeds, and then you can interpolate between them.

01:28:40.000 --> 01:28:47.000
And given that this was the only available fragility curve at that time for a stationary train,

01:28:47.000 --> 01:28:54.000
What I did is I mainly added equivalent shaking due to train speed on the x-axis.

01:28:54.000 --> 01:28:58.000
Mainly, you know, added, uh…

01:28:58.000 --> 01:29:03.000
Additional shaking due to train speed on top of the ground shaking due to the earthquake.

01:29:03.000 --> 01:29:07.000
And that… that's how… what allowed…

01:29:07.000 --> 01:29:14.000
us to use that stationary train mainly for different speeds.

01:29:14.000 --> 01:29:24.000
Okay, thank you.

01:29:24.000 --> 01:29:30.000
Okay, um… Brian, I wanted to know a little bit more about.

01:29:30.000 --> 01:29:37.000
functional recovery development of those types of codes, those types of ways of.

01:29:37.000 --> 01:29:46.000
giving instruction and advice to communities. What do you think is going to move the needle on communities' abilities to adopt the kind of code.

01:29:46.000 --> 01:29:53.000
improvements that… or even just… perceive and see the opportunity in.

01:29:53.000 --> 01:29:59.000
moving to a functional recovery-based. way of thinking about policy.

01:29:59.000 --> 01:30:04.000
So, where… where… what's the next step after developing the recommendations? How do we get there?

01:30:04.000 --> 01:30:09.000
Yeah, I think it boils down to one key part, and I think that's education.

01:30:09.000 --> 01:30:18.000
Um, we in the report to Congress, recognized one of the key recommendations was we need to educate our communities to help them know what they're getting out of their current building codes.

01:30:18.000 --> 01:30:25.000
Much like we need to educate our communities of what they have in their existing building stock that may already be vulnerabilities and risks.

01:30:25.000 --> 01:30:30.000
to their recovery. And so, the more we can educate the communities to help them understand

01:30:30.000 --> 01:30:35.000
That things like grocery stores, things like pharmacies, things like schools maybe even,

01:30:35.000 --> 01:30:38.000
will be down, and uh…

01:30:38.000 --> 01:30:43.000
without function for a significant amount of time, 6 months or more after a major earthquake.

01:30:43.000 --> 01:30:51.000
I think that will start to motivate them. There is some more work on the non-engineering side of things, in terms of some of the cost-benefit studies, in terms of long-term

01:30:51.000 --> 01:30:53.000
You know, to help those community leaders.

01:30:53.000 --> 01:30:58.000
Make those decisions and be able to justify those decisions, and I think a lot of the work you've heard about today

01:30:58.000 --> 01:31:02.000
Helps to contribute to that, whether it's the in-core studies or otherwise.

01:31:02.000 --> 01:31:07.000
And I also know that our Knee Hurt partners like NIST in particular,

01:31:07.000 --> 01:31:12.000
are eager to help us fill some of those gaps with kind of those cost-benefit studies, and also

01:31:12.000 --> 01:31:20.000
additional tools to help communities know where should they prioritize their time and energy in terms of which types of facilities they need to have up and running.

01:31:20.000 --> 01:31:24.000
More quickly, uh, in order to help them navigate this.

01:31:24.000 --> 01:31:31.000
Probably a staged adoption, I would think, as they kind of phase us in over time, of starting with their most critical needs,

01:31:31.000 --> 01:31:39.000
And then working it through their entire community.

01:31:39.000 --> 01:31:44.000
Oh, I lost your audio, Cheryl, sorry.

01:31:44.000 --> 01:31:45.000
Oh, there you are.

01:31:45.000 --> 01:31:46.000
You're muted again.

01:31:46.000 --> 01:32:03.000
I'm muted, I swear. That's… useful, and it reminds me a lot of David's work around the Bay Area. Like, how do you get policy spread? It's a lot about the awareness and being able to see models, and being able to see that it works, even if you didn't have the money that flowed in when you thought.

01:32:03.000 --> 01:32:08.000
Um, I'm kind of seeing from this panel. something that's long been of interest.

01:32:08.000 --> 01:32:20.000
Which is just kind of trying to… get alignment between engineering solutions or strategies, the money to implement them, and understanding of the benefits.

01:32:20.000 --> 01:32:30.000
You know, they all kind of have to line up and be calibrated in relation to one another, so the cost-benefit answers are really important to receptivity.

01:32:30.000 --> 01:32:39.000
to the solutions, um… You know, maybe you can design a solution that's a little cheaper and you don't need as much money to implement it, like.

01:32:39.000 --> 01:32:44.000
there's give in all of those, um, categories, but.

01:32:44.000 --> 01:32:45.000
I have to find the alignment, and it's not easy.

01:32:45.000 --> 01:32:50.000
Yeah, Cheryl, I'll just add in real quick on that. You know, I think a lot of…

01:32:50.000 --> 01:32:55.000
co-development over the years, and taking something voluntary to mandatory

01:32:55.000 --> 01:32:58.000
has benefited by…

01:32:58.000 --> 01:33:05.000
leadership, from example, or leading by example. And, you know, in many cases, like David said in the chat,

01:33:05.000 --> 01:33:10.000
It doesn't take that much more money to get this better-for-poor performance. It just needs to be more deliberate.

01:33:10.000 --> 01:33:20.000
I'm grateful the state of Washington right now has a little bit of a pilot program looking at our functional recovery provisions and how they might be able to apply them to their communities in the state of Washington.

01:33:20.000 --> 01:33:31.000
If you look back into, you know, sustainable design provisions, some of the green building code, uh, ideas got adopted and got more traction by the state of California using them for their own buildings.

01:33:31.000 --> 01:33:37.000
And so I think something similar to that is going to take this a long ways in helping others understand

01:33:37.000 --> 01:33:44.000
There's not a lot of mystery here. It doesn't take that much more. It's just a little bit more of a proof of concept.

01:33:44.000 --> 01:33:49.000
Well, I'm gonna take my bully pulpit at the end to say something I'd like to challenge.

01:33:49.000 --> 01:33:55.000
This community, the communities that we circulate in as earthquake, CFI advocates and professionals.

01:33:55.000 --> 01:34:01.000
Uh, we need to make. those success stories and.

01:34:01.000 --> 01:34:06.000
Buildings that got designed right, and the ones we know the costs and benefits of.

01:34:06.000 --> 01:34:13.000
We need to make that data really. Consolidated, standardized, transparent, and shareable, because.

01:34:13.000 --> 01:34:19.000
The ERI community has had a lot of difficulty this over this last year, trying to.

01:34:19.000 --> 01:34:25.000
get the stories of what we're not. getting anymore. The tragedy of what we're not.

01:34:25.000 --> 01:34:28.000
getting anymore, how to make that clear to people.

01:34:28.000 --> 01:34:33.000
And we just don't have enough. ready to tell stories.

01:34:33.000 --> 01:34:41.000
About the wins and the winning strategies. So even when those things are being threatened, it's been hard to communicate.

01:34:41.000 --> 01:34:49.000
the loss that we're experiencing over the past year, so… codify the information, follow David's example.

01:34:49.000 --> 01:34:58.000
Um, I want to applaud everybody who's contributed on this panel, asked great questions. This whole day has been super interesting, a lot of information flow.

01:34:58.000 --> 01:35:03.000
Alan, if you want to say anything else, uh… I think this is it. We're panning back over to sign off.

01:35:03.000 --> 01:35:08.000
Yeah. Yeah, we already passed 4PM.

01:35:08.000 --> 01:35:14.000
So, I want to thank our four incredible panelists, again, for their wonderful talks and discussion.

01:35:14.000 --> 01:35:17.000
And I'll pass it back to Sarah.

01:35:17.000 --> 01:35:24.000
And I, in turn, want to thank our wonderful moderators, Alan and Cheryl! Woo!

01:35:24.000 --> 01:35:25.000
So, the earth is always changing.

01:35:25.000 --> 01:35:29.000
Well, this was fantastic. This was a wonderful first day. Remember, there are 2 more days coming, so everybody please have a wonderful rest of today.

01:35:29.000 --> 01:35:34.000
Mountains and pallias rearranging.

01:35:34.000 --> 01:35:39.000
That's because the Earth is covered up with giant plates.

01:35:39.000 --> 01:35:40.000
And when they rub together, we get earthquakes.

01:35:40.000 --> 01:35:44.000
And come back when we start at. 9.30 Pacific tomorrow, unless… You are one of the Thunder Tacos in the opening session, which is a really long list of people, in which case, please come at least 9.15.

01:35:44.000 --> 01:35:45.000
So that we can get you all set up before the session begins.

01:35:45.000 --> 01:35:47.000
Shake, shake, shake, shake it, baby, the ground looks solid.

01:35:47.000 --> 01:35:50.000
So have a great evening, or night, or whatever time it is where you are, and see you tomorrow!

01:35:50.000 --> 01:35:57.000
But I've got