RC 結構耐震補強技術於建物生命週期之永續與韌性評估

Abstract

實現抗震韌性與永續性已成為當代建築設計與耐震補強的核心目標。一般來說，兩項目標往往需仰賴顯著的前期投資，導致在防災初期成本與長期效益之間存在取捨。因此，如何於評估階段即掌握各補強方案在不同面向上的成本效益，成為實務與研究上的重要課題。鑒於目前業界可選用之補強方案多元，為深入比較其在經濟、環境、社會與抗震韌性四大層面上的表現差異，本研究提出一套建築生命週期導向之多準則決策分析架構，從全生命週期角度評估各種結構補強方案間之效益權衡。依據美國聯邦緊急事務管理署（FEMA）所發布之 P-58 指南，分別採用遠域（Far-Field）與近斷層（Near-Fault）兩類地震歷時資料，進行耐震性能評估。本研究提出一基於數值積分的預期年化損失（Expected Annual Loss, EAL）計算方法，量化補強前後之風險降低效益；此外，本研究模擬山腳斷層錯動發生規模 7.3 且具速度脈衝特性之近斷層地震情境，以評估各補強策略在極端地震事件下之抗震韌性表現。在經濟面向上，為快速且準確地預估不同補強策略所需之初期投入成本，本研究建構一基於貝氏類神經網路（Bayesian Neural Network, BNN）模型，除提供準確金費預測外，亦同時量化其預測不確定性，以反映資料稀缺與物價波動下之潛在風險。進一步將震損減少效益與初期投入成本進行整合，分別計算經濟、環境、社會三個面向上的成本效益比（Benefit-Cost Ratio, BCR），作為多準則決策中之量化評估指標。最後，本研究將所建構之分析架構應用於一棟位於臺北之既有鋼筋混凝土建築進行案例驗證。結果顯示，該架構不僅具備良好之實務可行性，亦能與現行臺灣耐震評估流程高度整合，並拓展其評估範疇與應用深度。本研究所提出之整合性多準則評估架構可依據決策者之偏好進行調整，並全面涵蓋耐震補強策略於永續性與抗震韌性等多重目標下之效益表現，有助於提升補強決策過程之科學性與透明度，亦能為政策制定提供更具系統性與量化依據之支撐。

Achieving both seismic resilience and sustainability has emerged as a critical objective in contemporary building design and retrofit practice. However, these goals often entail substantial upfront investment, creating trade-offs between initial disaster-prevention costs and long-term benefits. To address this challenge, a life cycle based multi-criteria decision making framework is proposed for systematically evaluating retrofit strategies across four key dimensions including economy, environment, society, and seismic resilience. Based on FEMA P-58 guidelines, the framework incorporates both Far-Field (FF) and Near-Fault (NF) earthquake ground motion records to perform the seismic performance assessments. A integral based approach is developed to estimate the expected annual loss (EAL) more reasonable, enabling quantitative evaluation of risk-reduction benefits before and after retrofit. Additionally, a near-fault earthquake scenario with magnitude 7.3 and velocity-pulse characteristics at the Shan-Chiao fault is simulated to assess the resilience index of each strategy under extreme seismic events. From an economic perspective, a Bayesian Neural Network (BNN) model is developed to forecast initial retrofit costs with quantified uncertainty, accounting for risks stemming from data scarcity and market fluctuations. Combine the initial investment of retrofit with seismic loss-reduction benefits to calculate the benefit-cost ratio (BCR) as a key evaluation index. The proposed framework is applied to a real-world retrofit case involving a reinforced concrete building in Taipei demonstrates its practical feasibility, alignment with Taiwan seismic assessment process, and capacity to deliver a transparent, adaptable decision-support tool. Ultimately, the proposed approach enhances the evidence base for policy formulation and supports more robust, stakeholder-informed seismic retrofit decisions.