Blog

Earthquake Risk of Existing Buildings in Başakşehir: Findings from an ELER–HAZUS Pilot Study

Scenario damage estimation with ELER–HAZUS in two Başakşehir pilot regions and across the district: no heavy damage in the pilot regions, damage varies by building type.

October 9, 2025
Earthquake risk in Başakşehir: ELER–HAZUS pilot analysis cover image

This article summarises research based on Volkan Ergen’s master’s thesis (Gebze Technical University, 2017), published by Ergen, Zülfikar and Kılıç in Karadeniz Fen Bilimleri Dergisi (2025, 15(3), 1133–1163): The Evaluation of Earthquake Risk of Existing Buildings in Başakşehir District of Istanbul Province.

Why this study?

Evaluating the seismic risk of buildings in Istanbul’s newer residential areas matters for public safety and sustainable urban growth. The study selected two pilot regions in Başakşehir (Kayabaşı and Bahçeşehir 2. Kısım) and ran scenario earthquake hazard and damage estimation both for the pilot regions and for the whole district.

Method in brief

  1. Building inventory update: field surveys and building permit data were used to update the inventory for the pilot regions and for the whole district, and the data were transferred to QGIS.
  2. Analysis tool: the Earthquake Loss Estimation Routine (ELER) was used with the HAZUS building classification.
  3. Scenario earthquakes: four historical earthquakes and one instrumental earthquake affecting the Marmara region were modelled: 1509 Istanbul, 1766 Marmara, 1754 İzmit, 1894 İzmit and 1999 İzmit.

About the pilot regions

The building stock is mostly regular-form cast-in-place reinforced concrete: frame, shear wall and dual systems. In Kayabaşı, of 164 buildings, 59.76% have 8 or more storeys, 14.63% have 4–7 storeys and 25.61% have 1–3 storeys.

Main findings

  • Pilot regions: no heavy damage was estimated in the pilot regions under any scenario. In Kayabaşı, moderate damage appears in the 1766 Marmara, 1754 İzmit and 1509 Istanbul scenarios and slight damage in every scenario, with the highest numbers in the 1766 scenario. The second pilot region, Bahçeşehir 2. Kısım, gave very similar results. For the 1894 and 1999 İzmit scenarios the effect on the pilot regions is lower because the epicentres are farther away.
  • By building type (district-wide): type C3L2 (1–3 storeys, older code periods) has higher heavy-damage ratios than other types: 7.87% in the 1766 Marmara scenario and 5.16% in the 1509 Istanbul scenario. Type C2M4 (4–7 storeys, shear walls, newer code) shows no heavy damage in any scenario.

What the results indicate

  • Age, structural system and height: damage estimates are sensitive to a building’s age (code period), structural system and number of storeys. Mid-rise shear-wall buildings built to newer regulations show a better profile.
  • Planning: an updated district-scale inventory is a solid basis for prioritising urban transformation, for probabilistic damage studies and for comparison with other software.

Limits

The study uses deterministic scenarios. With the updated inventory, comparison against probabilistic approaches (different ground-motion models and return periods) is recommended. The results estimate damage distributions for building classes; they say nothing about the safety of any individual building, which requires its own engineering assessment.

StructHEALTH’s view on monitoring

This section is StructHEALTH’s own commentary, not a finding of the study. Scenario-based estimates can help decide where to look first. For selected critical buildings, monitoring with acceleration sensors adds measured data on how the structure actually behaves, which engineers can use alongside analytical predictions. See our solutions.