Why were so many buildings destroyed?
What should we do to avoid dying in the next earthquake?
Translated from my original Turkish post. I used Google translator then I fine tuned it in a haste. Hopefully, it makes sense as much as the original does.
-+-+-+-+
The discussions I watched after the earthquake were focused on punishing those who did not practice what we know. It is not asked how valid our knowledge is. Reluctantly, I decided to write this article. I say reluctantly because it is an area I know little about. Nevertheless, I believed someone should start the discussion so I decided to be the village idiot. I hope this article will be shared and the subject will enter the agendas of more competent and authoritative people.
-+-+-+-+
All those workshops, changed legislation, updated risk maps were in vain. We didn't go a long way. The Istanbul-Izmit earthquake in 1999 was supposed to be a milestone. Nothing has changed. The debates are the same, but the political identities of the speakers are different. Former opponents are now in power, and what they called murder then, they call fate today.
According to some studies, it is the fate of poor nations to die in an earthquake: for example, Li, Wang et al. (2021). In my opinion, the income level of the country is not a sufficient criterion to explain everything. While deaths exceeded 40 thousand in the last 7.7 earthquake in Turkey, only 520 died in an earthquake of 8.8 strength, 32 times higher than the last Maraş earthquake, in Chile in 2010, most of them with the post-earthquake tsunami (NEHRP Report, 2012). Both earthquakes occurred in the morning: 3:34 am in Chile, 4:17 am in Turkey. The population densities are similar: the population density of the earthquake-hit Maule region in Chile is around 50 per square kilometer; just like Maraş Malatya. Chile is not a very rich country, its national income is 317 billion dollars, and Turkey's national income is 819 billion dollars.
It is difficult to excuse the colossal neglect in Turkey with financial impossibilities. Everyone compares it to Japan; and the governments do not mind it. I benchmark it against Chile for a more realistic comparison. The situation is clear: we are far behind not only Japan but also Chile when it comes to protecting ourselves against earthquakes. I cannot explain why this is so using economic, social and geological parameters.
Turkey is doing something wrong, very wrong. The situation cannot be explained by corrupt municipalities alone or thieving contractors adding sea sand to cement. I don't think we have been able to establish an correct consistent system. The tools we use are either wrong or incomplete. No matter how much we declare new milestones after each disaster, we will continue to get the same sad results as long as we continue to make the same mistakes.
-+-+-+-+
As with illness, in an earthquake it is more effective to prevent a calamity than to try to cure its consequences. For a week, we watched how people in both state and voluntary institutions work selflessly despite all their planning and equipment deficiencies. It is not enough.
Buildings should not have been demolished and would not have been demolished if the right precautions had been taken. Now they are trying to catch the offending contractors. Very late. As Nasreddin Hodja said, if you don't want the water jug to be broken, you will throw the slap before the jug is broken. Let alone smacking, our state has encouraged faulty constructions by issuing amnesties to offending builders.
It is not impossible to build earthquake resistant buildings. There are many methods developed over the last century. For example, the isolation of the building foundation and the body from each other. Of course, it has a cost, but it's nothing compared to the lives gone. Japanese construction firm Nice Corporation says the seven-story building with basic containment costs 13 to 15 percent more. In the video below, you can see how a hospital staff in Japan survived the 2011 Tohoku earthquake with a magnitude of 8.9 without a nosebleed. Tables and chairs slide left and right with the rocking building, but the carrier structure is not damaged, even the plasters do not crack.
While watching the video above, you will notice the calmness of the people. As the hospital shakes, doctors plan how to organize for first aid to the injured who are expected to arrive in the emergency room after the earthquake is over.
This system has now started to be used in Turkey as well. This video shows the stretch on the foundations of Malatya Training Hospital:
As I learned from Ayşe Hortacsu's LinkedIn message, the new Maraş hospital will use a similar set-up and Dörtyol Hospital was not demolished because it used this system. So our problem is not a lack of technology access.
-+-+-+-+
Why were the buildings destroyed?
Not only the old but also the new buildings were destroyed. For example, the two pictures below are of a building in Hatay before and after the earthquake.
The situation in the surrounding buildings is even worse, but there are also those that did not collapse:
The construction of new buildings in accordance with the legislation is a legal obligation. If they are still falling apart, it can be explained in two ways:
The legislation, namely the Turkish Earthquake Code (TDY), is incorrect or insufficient; or
Contractors do not comply with the legislation, so there is an audit error.
The second proposition is definitely correct. We hear that there are many contractors who do not comply with the legislation. Unsupervised buildings do not only comply with earthquake codes, but also with general building codes. Since there are so many such buildings, it has not been possible to test the assumptions and the methodology used in the Turkish Earthquake Code itself until now. But you have to start somewhere.
-+-+-+-+
Turkish Earthquake Legislation
Türkiye Bina Deprem YönetmeliğiThe Turkish Building Earthquake Code (TDY),has been revised 7 times in total, including 1947, 1953, 1961, 1968, 1975, 1998, 2007, and 2018, which is still in effect.
-+-+-+-+
Preparation process
Wide participation is required in the preparation of regulations concerning all areas of life, especially those who are expected to comply with this legislation. There are two reasons for this. First, many issues that academics and bureaucrats may not think of can be brought up by companies. Second, involving companies in the process makes it easier for them to take ownership of the outcome.
Looking at the official documents, it seems that Turkey's Earthquake Planning is entirely in the hands of academics and bureaucrats. For example, the head of the Monitoring and Evaluation Board, which started the National Earthquake Strategy and Action Plan (UDSEP-2023) ten years ago, is chaired by Dr Fuat Oktay, one of the chief advisors to the President, and three of the other six members are professors and two are bureaucrats; I do not know the profession of the seventh member who was there as the president of the Turkish Earthquake Engineering Association.
In the 10th annual report of UDSEP published last year (2022), among the 97 people listed as Advisory Board and commission members, there is only one person who is not a civil servant or academic, and that is Bülent Atamer, who is named as the representative of the Turkish Contractors Association.
However, I think it is essential for the industry to be involved in the process from day one for the right product and widespread acceptance.
-+-+-+-+
Outreach and stakeholder training
The construction industry needs courses, manuals to digest a regulation that they were not involved in the preparation process. Course material containing sample calculations and case studies should be prepared for university professors. Also, earthquake code should be included in general civil engineering software packages.
Responsibility for doing these is clearly given to AFAD in Articles 12 and 17 of the Law on the Organization and Duties of the Disaster and Emergency Management Presidency.2
I searched the AFAD website carefully and couldn't find anything about my announcement and training. There is the Turkish Journal of Earthquake Research and this journal could perhaps be thought of as a discussion forum on earthquake risk and earthquake regulation, but it looks like any scientific journal as it stands. Its difference from its international counterparts is that it is Turkish. Having a magazine like this is better than none, but it's not enough.
-+-+-+-+
Earthquake Education and Research Center
There is an urgent need for a research center that is organized independently of state authorities but includes civil servants, academics and industry experts. An income base with government support and subscription component can be created. Similar organizations in other countries can be taken as an example. The minimum mission of such a center should include:
Inform the public;
To train industry workers at appropriate levels;
Software tools;
Comparison against other country legislation; and
To make recommendations to relevant government agencies.
-+-+-+-+
Technical problems in Earthquake Legislation
As a Mechanical Engineer, I worked for many years on the design and analysis of steel structures used in large machines. I can't do reinforced concrete calculations, but I think I have enough engineering knowledge and experience to read and analyze the load determination method that a building can be exposed to. Read what I'm about to write below with that eye.
While defining the earthquake risk in the Turkish Earthquake Code, a mixed approach of Eurocode 8 and US legislation was followed. US and European standards are quite different from each other in terms of risk analysis. When combining two different methodologies, it is necessary to explain very well why some choices are made the way they are. Otherwise, a product may emerge that is not self-consistent. I'm not saying it is inconsistent yet because I didn't have time to research it in detail while preparing this article. But there is a danger of it happening.
Eurocode 8, used in EU countries, bases earthquake loading on the peak ground surface acceleration (PGA) that will occur on rocky soils. This is the parameter shown on the map presented on the AFAD website as the Turkey Earthquake Hazard Map:
AFAD has posted the PGA map on its website, but unlike EU countries, an acceleration spectrum model with two coefficients is used in TDY earthquake loading calculations, not PGA:
he method used in TDY resembles ASCE 7-16 used in the USA before 20223. However, there is a big difference between the two in terms of the definition of SDS and SD1 values. The American standard uses values with a 2% probability of being exceeded in 50 years. In the Turkish Earthquake Code, there are four earthquake level categories: DD-1, DD-2, DD-3 and DD-4. These categories represent the probabilities that earthquake risk parameter values may occur in fifty years:
DD-1 : 2% chance of exceeding in fifty years
DD-2 : 10% chance of being exceeded in fifty years
DD-3 : 50% chance of being exceeded in fifty years
DD-4 : 68% chance of exceeding in fifty years
Of course, as the probability of exceedance increases in fifty years, the level of design earthquake load decreases. While 2%, i.e. DD-1 conditions, are obligatory in the USA ASCE-7 16 standard, flexibility is given to the designer engineer in TDY. For tall buildings (over 70 m, residential building), the designer working according to TDY rules can choose one of four levels. The performance target changes according to the level he chooses. If the designer chooses DD-1, that is, the highest seismic load, the performance goal is to prevent collapse (PI). If the engineer designing the tall building chooses DD-4, that is, the lowest earthquake load, then the performance target is Uninterrupted Use (CC). Whether the building performance target has been met or not is assessed by computing the stresses and deformation. For buildings below 70-metre high, the only option is DD-2 and the performance target is Controlled Damage (HR).
In order to show that there are significant differences between the results of the preferences given by TDY to the designer engineer, I calculated the design loading that a 40-m high apartment building in Istanbul and Antakya would be exposed to in an earthquake according to the USA (DD-1) and TDY-KH(DD-2) conditions. :
In this graph, the x-axis is the building's natural period ( T ); the y-axis is the horizontal shaking acceleration of the building foundations. The natural period is a feature of the building. The opposite of the concept of natural frequency ( f ) , which is widely used in vibration calculations in mechanical engineering The literature says that typical natural periods are between T =0.4 and T= 2 but there may be very different numbers depending on the design of the building. Returning to the graph above, for typical T values ( T≥ 0.4 s), the earthquake loading difference between Istanbul and Antakya is negligible if DD-2 is accepted, but there are significant differences for DD-1. In other words, if you had to comply with the American ASCE 7-16 standard, in which DD-1 is mandatory, you would have to produce different building designs for Istanbul and Antakya, or you would have to design it for Antakya and design the same design for Istanbul, knowing that the building in Istanbul would be more durable (and maybe more expensive) than necessary. Using DD-2 (ie per TDY) removes this dilemma.
The idea of different earthquake levels may have been taken from Eurocode 8. Eurocode 8 stipulates that the earthquake loading that should be used for anti-collapse designs be accepted based on the reference response time of seismic action (TNCR), or equivalently, the reference probability of exceedance in 50 years (PCNR).
These are very technical issues and I don't want to go further. Eurocode 8 recommends the following values for reference
For example, Italy uses Eurocode reference values for structures with a design life of 50 years, tightening the requirements for buildings that are expected to survive longer (TCNR=712.5 years (PCNR=7%) for buildings of 75 years); TCNR=950 years for a design life of 100 years. , PCNR=5%). England, which has a very low earthquake risk, has accepted TCNR=2500 years (our DD-1 means). I haven't looked it up specifically, but the earthquake risk in England is low, and an earthquake expected every 2500 years may be less intense than the earthquake that is expected to repeat every 100 years in Italy.
Comparing the Turkish Earthquake Code with the standards used in other countries is an urgent requirement, but it is a project that should be carried out jointly by local and international experts who are more knowledgeable than me. For now, I will only observe that the Turkish Earthquake Code is an eclectic model, and therefore, like all eclectic models, inconsistencies may arise if care is not taken. The only scientific article I could find about civil engineering admissions in the last TDY is Sucuoğlu (2019), which was published in AFAD's Turkish Earthquake Research Journal in 2019. It's a well-written article, but the researcher limited himself to summarizing TDY items; He did not touch upon the comparative justification of assumptions.
I also read two Master's theses comparing the Turkish Earthquake Code with Eurocode 8: Tekince (2015) and Kazancı (2018). They came to conflicting conclusions. One found TDY more conservative, the other Eurocode 8. Such a contradiction between two scientific studies bothered me.
Finally, in all countries, the scope of earthquake codes is to estimate earthquake intensity and convert it into load vectors to act on the building; and to define the building geometries that are expected to be the most earthquake resistant. The detailed design and construction of the building is carried out in accordance with the current general construction laws and regulations. There are too many buildings in Turkey that do not comply with the general construction procedure, let alone the earthquake code requirements. Therefore, it can be said hat we succeed in applying the existing rules, and then think better of those rules . However, both can be done together.
-+-+-+-+
Conclusion
I have suggested a few things that I believe should be done immediately in the flow of the article above. To draw more attention to them, I list them here at the end of the article:
Remove outreach, public and stakeholder education duties from AFAD's mission, and assign them to a new research center organized independently from state authorities but includes civil servants, academics and sector experts. An income base with government support and subscription component can be created. Similar institutions in other countries can serve as examples. Earthquake taxes can be used to implement policies that this institution will create independently of governments.
A study that compares the Turkish Earthquake Code and the US (and maybe European) standards, explains the differences, and proposes changes in the TDY if necessary, should be conducted with the participation of international and local experts immediately. As an example, take a look at the study report published after the Chile 2010 earthquake (NIST Report, 2012).
Again, immediately and while the memories are still fresh, the following classification should be made for each city separately:
How many of the destroyed buildings comply with the general zoning and construction legislation?
What is the number of buildings that are in compliance with the general zoning and construction legislation but not in compliance with the earthquake regulation?
Is there any building that was destroyed even though it is in compliance with both the general zoning and construction legislation and the earthquake regulation? What is the rate?
These are studies that will take time, but they can be carried out as graduate and doctoral theses in our universities.
._._._._
References
Building Seismic Safety Council (2010) 2020 NEHRP (National Earthquake Hazards Reduction Program) Recommended Seismic Provisions: Design Examples.
Charney, F. A., Heausler, T. F., & Marshall, J. D. (2020). Seismic Loads Guide to the Seismic Load Provisions of ASCE 7-16. American Society of Civil Engineers.
Global Earthquake Model (GEM). Global Seismic Hazard Map. https://cloud-storage.globalquakemodel.org/public/Global%20Maps/version_2018.1_20200928.png
Harris, J L, Speicher, M S (Şubat 2015). Assessment of First Generation Performance- Based Seismic Design Methods for New Steel Buildings, Volume 1: Special Moment Frames. NIST Technical Note 1863-1.
Harris, J L, Speicher, M S (Şubat 2015). Assessment of First Generation Performance- Based Seismic Design Methods for New Steel Buildings, Volume 1: Special Concentrically Braced Frames. NIST Technical Note 1863-2.
Harris, J L, Speicher, M S (Şubat 2015). Assessment of First Generation Performance- Based Seismic Design Methods for New Steel Buildings, Volume 3: Eccentrically Braced Frames. NIST Technical Note 1863-3.
Kazancı, S. (2018). Türki̇ye Deprem Yönetmeli̇ği̇ 2007 Ve Eurocode 8’e Göre Tasarlanan Betonarme Binalarda Doğrusal Analiz Yöntemlerinin Karşılaştırılması, İstanbul Gelişim Üniversitesi Fen Bilimleri Enstitüsü.Kazanci, S. (2018).
Li, Y., Wang, Y., Zhang, Y., Zhou, X., & Sun, H. (2021). Impact of economic development levels on the mortality rates of Asian earthquakes. International Journal of Disaster Risk Reduction, 62, 102409. https://doi.org/10.1016/j.ijdrr.2021.102409
NIST Raporu (2012). Comparison of U.S. and Chilean Building Code Requirements and Seismic Design Practice 1985–2010.
Özmen, B.(2021). Ozmen, B.(2021). TOBB sunumu:TOBB presentation: ‘İşletmeler, Yerel Odalar ve Borsalar için Deprem Mevzuatı’
Sesetyan, K., Demircioglu, M.B., Duman, T.Y. et al. A probabilistic seismic hazard assessment for the Turkish territory—part I: the area source model. Bull Earthquake Eng 16, 3367–3397 (2018). https://doi-org.ezproxy.library.uq.edu.au/10.1007/s10518-016-0005-6
Sucuoğlu, H. (2019). New Improvements in the 2019 Building Earthquake Code of Turkey. Turkish Journal of Earthquake Research 1 (1), 63-75, June 2019.
Tekince, Ö. (2015). Betonarme Binalarda Doğrusal Analiz Yöntemlerinin TDY 2007 ve EC 8’e Göre karşılaştırılması, İTÜ Fen Bilimleri Enstitüsü.