Hello guys,

Amongst some common questions which I have found that people ask about evacuation modelling and particularly about the evacuation simulation models' capability to represent real situations is:

- In Earthquakes situations, is it possible to model occupants' movement taking into account the vibrations during this natural phenomenon?

For instance, do you know any model which try to represent this vibration into the simulation when the occupants are trying to escape within an enclosure?

This question also came strongly to my mind after these terrible earthquakes which happened more recently in Asia.

cheers,

Rodrigo

Hey Rod,

I don't know of any models which can represent the affects of earthquakes on pedestrian behaviour- maybe recent events in China should prompt us to start developing them?

The nearest type of behaviour I could think of would be with regards to the movement of pedestrians on ships using variable heel and trim, however, this does not seem that similar a "shaking" affect - I'm guessing many models which already include the affects of heel and trim could accommodate earthquake movement. The problem is, we don't have any data (to my knowledge) on how pedestrians behave....sad as it may sound, maybe there would be a chance of collecting some from the earthquake zone in China?

Hi Michael,

That's a good point. I also thought about this, once I believe that the maritimeEXODUS, for instance, simulates this movement in ships. But as you mentioned, despite of this being the closest approach to an earthquake, this is quite different.

This might be a good start point of a research though (collecting such type of data). I would assume that in Japan, we might find some data (but maybe not).

But this for sure, is something that we should starting to look, once it seems that our lovely planet Earth is starting to shake even more often everywhere. Two weeks ago, we had also a small earthquake in the city of Sao Paulo in Brazil. The earthquake reached 5.2 in the Richter scale (it is considered a small earthquake; but people needed to evacuate and small vibrations were felt and even these small vibrations do impact people's movement. Besides that, I believe that also the psychological factor might play a great importance there, once people will have this on their minds: "the building is gonna collapse, I need to escape from here..."). Cities like Mexico City, some cities in the Southwest of the USA and several parts of Asia have historical events involving earthquakes.

I believe that we have two issues to be understood in terms of evacuation on earthquakes:

1) how people react/behaviour during earthquakes not only in terms of movement but even in terms of psychology. But I would restrict this at first place to only the physical movement, since the human behaviour aspect can be just an extension of any other fatal risky event (i.e., fire, explosion, flooding etc.).

2) to estimate the RSET taking into consideration the ASET. But the ASET is now analyzed in terms of how long the earthquake's effect on the building structure would take to collapse it. So, similarly to a fire event, instead of predicting the time to reach the flashover phenomenon, we need to have an idea of the time to reach the collapsing point and besides that how the vibrations on the structure would impact the occupants' movement.

I believe that it's an interesting research and more important it's needed. I think at first moment, we should try to collect data, as you said, once there is no data (if there is, it's not publicly well known) on how people react and moves during an earthquake.

That's a good discussion though.

cheers,

Rodrigo

P.S.: If I find something about this issue, I'll post here.

Good ideas Rodrigo!

Thinking about the dynamic heel and trim data, a lot of this was collected from ship simulation rigs(e.g. SHEBA facility - http://www.shipevacuation.com/?/536/741 ). Now, I wonder if these same rigs could be used to test different grade earthquakes and collect pedestrian data??

I think the psychology of pedestrians in these situations is perhaps the most important issue to address if we ever want to simulate them (and so improve the survivability!). Perhaps an interview based survey is required of survivors of such events to asked people what they did and why?

Thanks for the link Michael. An interview associated with videos footage could ideal for collecting data and developing an understanding on this issue.

Ideally it would be getting these data (probably from Japanese researchers??). But anyway, this is definitely an interesting and relevant subject to research though.

cheers,

Rodrigo

These are some links which might be useful for a general understanding on Earthquakes:

http://www.eri.u-tokyo.ac.jp

http://www.japan-guide.com/e/e2116.html

http://www.seismo.unr.edu/ftp/pub/louie/class/100/effects-kobe.html

http://www.seinan-gu.ac.jp/~djohnson/natural/quakes.html

http://www.oes.ca.gov/Operational/OESHome.nsf...

http://www.timesonline.co.uk/tol/news/uk/article1719003.ece

Some of these links might not be that useful, but this is a starting point. I will try to find specific data on evacuation during earthquakes. I believe that some japanese research institute might work with this and also in California, USA.

cheers.

R

they is a few interesting videos on youtube of evacuations during earthquakes also:




there are loads more about.

yes, I have seen these and others. The first one, in my opinion, is not that good. But it's better than nothing. There are quite few videos, but I haven't seen that many videos regarding to evacuations during earthquakes in buildings. Maybe I need to search more and better though.

I thought these links might be useful though:


http://www.city.ota.tokyo.jp/english/emergencies/0603evacuation.html

http://www.city.shinagawa.tokyo.jp/s_foreign/english/earthquakes.html


And this other link brings a simple document on preparing and educating for "earthquake drills". It distinguishes the difference between fire drills and earthquake drills. For my understanding, it was interesting and helpful:

http://www.phivolcs.dost.gov.ph/news/eqdrill.pdf


regards,

Rodrigo

As I understood it, in Japan people are told not to try to keep standing on their feet, but to crawl. So, one would have to simulate crawlers. Crawling might be not that different during an earthquake than in a normal situation compared to the difference that exists for walking in both situations.

There is at least one empirical study I know of, that investigates crawlers' flow through a bottleneck:

Evacuation of crawlers and walkers from corridor through an exit
Ryoichi Nagaia, Masahiro Fukamachia and Takashi Nagatani
doi:10.1016/j.physa.2005.11.031
Physica A: Statistical Mechanics and its Applications
Volume 367, 15 July 2006, Pages 449-460

The other question is, of course, if people really start crawling and do not spend their time trying to keep standing on their feet.

There is also another study more recently which looked at the speed of crawlers:

"Development and validation of a crawling model in an existing computational
egress tool", Steve Gwynne, and R Muhdi, J Davis, Interflam 07.


The main problem I can see with crawling during an evacuation is the decrease speed and subsequent increase time needed to evacuate which would increase the chances of the individual being involved in an accident. Compounded by people having to traverse stairways in multi-level structures, I would think crawling is not such a good idea.

This research group from UCL, The EPICentre, "aims to bring together earthquake engineers, social scientists, coastal engineers and statisticians to provide guidance for key decision makers on where and how to invest resources in order to mitigate effectively for future earthquake losses":

http://www.epicentreonline.com

The advice given to me when I was a kid was to hide under a table or under a structural beam of the house/school/building when an earthquake would strike. Once the earthquake stopped we should evacuate. In this sense the earthquake acts as an alarm, the occupants perform some initial protective actions and then once the earthquake subsides the evacuation can starts. If the earthquake is very severe the story is very different! Note that earthquake motions are quite different from the heel/trim ship motions as it is more like a lateral shaking. Perhaps as an initial attempt to model this kind of scenarios the evacuation models should allow the modelling of the procedures employed in each structure/country. This would be the simplest thing to do. Modelling the effects of an earthquake in the case when there are structural damages is a different story. Points raised by Rodrigo further above in post 78 are also things that have to be considered.

http://www.cnn.com/2010/WORLD/americas/01/14/haiti.earthquake/index.html

http://g1.globo.com/Noticias/Mundo/0,,MUL1446514-5602,00.html

http://news.bbc.co.uk/2/hi/americas/8458439.stm

http://www.bbc.co.uk/portuguese/noticias/2010/01/100113_haiti_galeria_story.shtml

http://www.bbc.co.uk/portuguese/noticias/2010/01/100114_haiti_pessoas_ruas_rw.shtml

It seems a lot of the buildings collapsed during the earthquake (similar to that in the Chinese earthquakes which happened over a year ago) which would mean people wouldn't of had an opportunity to evacuate them. Does anyone know any standards or guidelines for how long buildings are intended to stay up during an earthquake of a given magnitude? I guess this would define the ASET.

Hi colleges,

I was reading your discussion about pedestrian modelling during earthquakes. In relation with how people behave in terms of psicology and response times, you can watch the following video about a earthquake in Caracas last year.

http://www.youtube.com/watch?v=Ij5FyHcTt0w

Regards

Thanks Arturo for the link. it is an interesting video to watch though.

cheers,

Rodrigo