Risk evaluation for earth-fill dams due to heavy rains by response surface method

Shiying Zheng, Shin ichi Nishimura, Takayuki Shuku, Toshifumi Shibata, Tsubasa Tateishi

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

This paper discusses a risk evaluation for earth-fill dams due to heavy rains. The detailed method employs a flood analysis and land use data to calculate the costs of the inundation damage in the downstream areas of earth-fill dams. The procedure to calculate the damage costs requires a lot of labour. Since a huge number of earth-fill dams exist in Japan, a straightforward method is needed. The response surface method, one of the surrogate models, is proposed in this study to reduce the calculation effort. The optimum response surface is firstly evaluated by cross validation, and then the accuracy is verified by comparing the damage costs obtained by the response surface method with those obtained by the detailed method for the earth-fill dam sites. To calculate the risks, it is necessary to determine the probability of overflow failure due to heavy rains. The risk of breaching is calculated from the product of the probability of overflow failure and the estimated damage costs. The accuracy of the response surface method is assessed by comparing the risk rankings of the dams, which is the priority in dam renovations, between the detailed and the response surface methods.

Original languageEnglish
Pages (from-to)572-585
Number of pages14
JournalGeorisk
Volume17
Issue number3
DOIs
Publication statusPublished - 2023

Keywords

  • Earth-fill dam
  • response surface method
  • risk evaluation

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Building and Construction
  • Safety, Risk, Reliability and Quality
  • Geotechnical Engineering and Engineering Geology
  • Geology

Fingerprint

Dive into the research topics of 'Risk evaluation for earth-fill dams due to heavy rains by response surface method'. Together they form a unique fingerprint.

Cite this