PhD Oral Exam - Bikram Patra Kesharee, Civil Engineering

When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.

Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.

Abstract

Dams are indispensable national assets that offer key benefits such as water storage, hydroelectric power generation, and flood control. However, these structures pose significant risks to downstream communities and infrastructure without appropriate maintenance and operation. The essentiality of rigorous management, consistent inspections, and adherence to modern safety standards cannot be overstated, aiming to preserve their functionality and ensure public safety. With a global count exceeding 61,000 large dams, the aging state of this infrastructure is a growing concern, particularly as many exceed 50 years of service and some approach or surpass a century. This aging, structural, operational, and budgetary shortcomings, and diverse safety assessment practices highlight the pressing need for enhanced dam safety measures. Obsolete seismic design criteria used during the initial construction of most dams, when compared to current engineering knowledge, further amplify the urgency for a thorough reevaluation of dam safety in line with today’s standards. Such reevaluation is crucial for improving operational efficiency and bolstering the structures' resilience to severe seismic activities. Although dam failures are relatively rare, their potential to cause catastrophic damage, underscored by over 300 significant global incidents, necessitates a focused and strategic approach to seismic safety assessments to effectively mitigate risks and ensure the continued provision of the benefits dams offer.

The safety assessment of dams is conducted through two primary methodologies: the Standard-Based Approach (SBA) and the Risk-Based Approach (RBA). The latter has gained favor for its holistic assessment of risks across individual dams or portfolios, enabling the prioritization of remedial actions and efficient resource allocation. RBA's widespread adoption is attributed to its ability to incorporate probabilistic risk elements comprehensively. Common tools for failure probability estimation in risk analysis include the analytical hierarchy process (AHP), fault tree analysis (FTA), and event tree analysis (ETA), although these tend to be unidirectional and may not fully account for event correlations. To address these limitations, the Bayesian method for risk analysis (BMRA) has emerged as an advanced alternative, offering the capacity to model complex systems through the integration of various data sources and facilitating an updatable knowledge base for accurate event probability analysis and risk-reliability scenarios.

The thesis is organized into four main sections, starting with a comparative analysis of existing safety assessment practices to identify their strengths and limitations. It then delves into three case studies focused on: (i) the selection of consistent numerical models, (ii) evaluating the impact of ground motion selection techniques and record-to-record variability on fragility and risk assessments, and (iii) assessing the effects of material property variations due to aging and construction heterogeneity on fragility. The study employs various damage indices, such as normalized crest displacement (NCD), damage area ratio (DAR), and tensile stress at critical dam locations, to develop fragility curves for a comprehensive safety evaluation. Based on these analyses, a new framework is proposed.

The thesis begins with an in-depth review of SBA and RBA, highlighting their limitations and introducing the Bayesian method as an improvement. It presents a reliability-based seismic safety assessment framework, incorporating both deterministic and probabilistic numerical simulation approaches. Given the challenges posed by complex finite element models in describing dam-foundation-reservoir systems and the extensive computational demands of probabilistic analyses, the study advocates for models of increasing complexity that balance efficiency with accuracy. This approach ensures the reliability of safety evaluations by adequately representing the physical behavior of dams under seismic loads.

In assessing seismic performance, the study covers linear and non-linear analyses of Koyna and Pine Flat dams, examining various model complexities and solution approaches to evaluate model variability. It highlights the importance of ground motion selection techniques, such as ASCE 7-16 and Conditional Mean Spectrum (CMS), in accurately representing seismic hazards and refining fragility assessments. A detailed fragility assessment for the Pine Flat DFR model, incorporating 110 ground motions selected through ASCE 7-16 and CMS methods, illustrates the effects of ground motion selection and record-to-record variability on dam safety evaluations. Additionally, the study explores the significant impact of material degradation due to aging and construction heterogeneity on the seismic fragility of concrete gravity dams, emphasizing the need for periodic reassessments.

The proposed framework for seismic risk assessment of concrete gravity dams integrates advanced ground motion selection techniques, progressive numerical modeling, and a comprehensive set of damage indices and considerations for material degradation. This approach not only quantifies the impact of ground motion variation and material degradation but also provides engineers and dam safety professionals with the necessary tools and insights for informed decision-making. By safeguarding dams against seismic threats, this framework aims to protect lives, ecosystems, and infrastructure from the potential consequences of dam failures, reflecting the latest advancements in the field.