Application of Risk Analysis for Dam Safety Assessment

Jong‐Seok LeeJournal of Korea Water Resources Association 2002 · DOI

Abstract

댐 안전성 평가의 주된 목적은 댐이 지니고 있는 잠재된 위험성을 미리 예측하고, 합리적인 방법으로 밝혀진 위험성을 줄여나가는 것이다. 본 연구에서는 위험도 해석기법을 이용하여 댐의 붕괴를 가져올 수 있는 각 상황별 댐의 위험성을 결정하였다. 또한, 댐의 위험성을 줄이기 위해 제시된 대안들에 대해서도 댐의 붕괴 위험성뿐 아니라 경제적인 측면에서의 비교, 평가를 실시하였다. 위험도 해석기법의 적용을 통해 댐의 안전을 책임지고 있는 관계자들은 댐의 위험성에 대한 구체적인 정보와 댐의 위험성을 개선할 수 있는 다양한 대안들에 대한 보다 넓은 이해를 가질 수 있다. 또한, 댐 정책 결정자들은 이러한 댐의 위험성에 대한 정보와 이해를 바탕으로 댐의 위험성을 효과적으로 줄여나갈 수 있는 대안들에 대해 자금과 자원을 집중하여 댐의 위험성을 효율적으로 관리할 수 있을 것으로 기대된다.

Introduction

Traditional dam safety evaluations often rely on a binary "safe or unsafe" approach based on deterministic design standards. However, as dams age and climate patterns shift—leading to updated estimates for Probable Maximum Floods (PMF)—many existing structures technically "fail" current standards. The challenge for engineers and policymakers is that resources for remediation are limited. This research addresses the gap between raw engineering data and rational economic decision-making in infrastructure management by answering how risky a dam is and which upgrade provides the most safety per dollar spent.

Methodology

The study adopts a four-stage risk analysis framework:

  1. Risk Identification: Identifying potential failure modes such as overtopping during floods, seismic instability, or internal erosion (piping).
  2. Risk Estimation: Using Event Tree Analysis (ETA) to map out the sequence of events from a "loading condition" (e.g., a specific flood level) to a "consequence" (e.g., dam breach and downstream damage).
  3. Risk Evaluation: Comparing the calculated risks against established safety thresholds, such as the US Bureau of Reclamation’s $10^{-4}$ per year failure probability.
  4. Risk Management: Evaluating 19 different structural and operational alternatives based on their Incremental Risk Cost and Benefit/Cost Ratio.
Event Tree Analysis (ETA) framework used for estimating failure probabilities across different loading conditions.

The study utilizes the National Weather Service DAMBRK model for flood wave simulation and integrates the DeKay and McClelland (1993) model to estimate potential loss of life (LOL).

Case Study: Alamo Dam

The primary case study is the Alamo Dam in Arizona, USA. The analysis incorporates hydrological data (Annual Exceedance Probabilities for various flood stages), seismic data (Peak Ground Acceleration probabilities), and economic damage functions using FEMA’s standard inundation depth-damage curves for residential, commercial, and manufacturing sectors.

Geographic location of Alamo Dam and Lake within the state of Arizona.

Results and Discussion

The research moves beyond viewing a dam as a single unit by deconstructing risk into specific loading conditions. For the Alamo Dam, flood risk was the primary driver ($1.89 \times 10^{-6}$/year), followed by seismic risk ($0.61 \times 10^{-6}$/year). The total failure probability was calculated at $4.67 \times 10^{-6}$ per year, which is within the acceptable range of international safety standards.

A key finding is the "Bang for the Buck" metric. The most effective technical solution is not always the most rational. While some massive engineering projects (like FER1) reduced risk to near zero, their cost-benefit ratio was poor compared to smaller, targeted interventions. For instance, FER1 cost $48.52M, whereas FR14 (installing a 5ft parapet wall) emerged as the most cost-effective solution, offering a high benefit-cost ratio for a relatively low investment of $2.61M.

Conclusion

Risk analysis provides a rational ground for decision-makers, allowing them to justify budget allocations with concrete data on risk reduction. While probabilistic models are sensitive to input data for rare events, they offer a superior alternative to purely deterministic mandates. Future work should focus on developing standardized risk-based safety criteria specifically for the Korean regulatory environment, incorporating localized depth-damage curves and regional hydrological models.

Summary updated

Abstract

댐 안전성 평가의 주된 목적은 댐이 지니고 있는 잠재된 위험성을 미리 예측하고, 합리적인 방법으로 밝혀진 위험성을 줄여나가는 것이다. 본 연구에서는 위험도 해석기법을 이용하여 댐의 붕괴를 가져올 수 있는 각 상황별 댐의 위험성을 결정하였다. 또한, 댐의 위험성을 줄이기 위해 제시된 대안들에 대해서도 댐의 붕괴 위험성뿐 아니라 경제적인 측면에서의 비교, 평가를 실시하였다. 위험도 해석기법의 적용을 통해 댐의 안전을 책임지고 있는 관계자들은 댐의 위험성에 대한 구체적인 정보와 댐의 위험성을 개선할 수 있는 다양한 대안들에 대한 보다 넓은 이해를 가질 수 있다. 또한, 댐 정책 결정자들은 이러한 댐의 위험성에 대한 정보와 이해를 바탕으로 댐의 위험성을 효과적으로 줄여나갈 수 있는 대안들에 대해 자금과 자원을 집중하여 댐의 위험성을 효율적으로 관리할 수 있을 것으로 기대된다. The main purposes of dam safety assessment are to predict risk of an existing dam and to reduce the identified risks reasonably. This study determined the risks of an existing dam for each different dam breach scenario using risk analysis. In addition, the application of risk analysis in the dam safety assessment made possible to compare and evaluate a variety of alternatives that may reduce risk of an existing dam with respect to the possibility of dam failure and economic efficiency, On the ground of the risk analysis results, the dam safety analysts can obtain better understanding and more information regarding the risk of and existing dam and recommended alternatives. Decision-makers will be able to manage risk of an existing dam efficiently by spending their affordable money and resources on the alternatives that are expected to reduce the risk of an existing dam practically.
BibTeX
@article{lee2002application,
  author = {Jong‐Seok Lee},
  title = {{Application of Risk Analysis for Dam Safety Assessment}},
  journal = {Journal of Korea Water Resources Association},
  year = {2002},
  doi = {10.3741/jkwra.2002.35.6.651},
}
Figure