非預期事件如意外事故和異常狀態的調查與報告，對潛在風險的辨識及控制極具關鍵性，同時還可符合製程安全管理的法令規定和OHSAS 18001要求。事故調查主要目的為了解事故發生的真相，進而改善風險控制與管理系統缺失之處，避免相同、相似的災害一再發生，而非著重於責任歸屬或指責特定人士。因此，深入探討意外事故的直接原因，並找尋事故的貢獻因子和基本因子是非常重要的。事故調查最常犯的錯誤是只調查事故的直接原因，而並未深入調查管理系統失效層面，儘管勞動檢查法第26條，已經在台灣執行超過15年，但是我國仍然缺乏有效的意外事故調查方法和專精事故調查的人員，也成為預防職業傷害和工安事故的主要障礙，因此本研究主要的目的為提供實用和有效的意外事故調查方法。 本研究以系統化的意外事故調查方式引領事故調查，一般而言，事故調查主要由證據收集、分析證據和根本原因分析幾個步驟所構成。本研究認為事件及成因圖在眾多事故調查技術中，為較具系統化和邏輯性的方法，可以協助證據與資料的彙整並且重新檢視資料不足之處，但由於事件及成因圖本身的定義較為籠統，因此本研究經由修正事件及成因序列技術，以提升該技術的應用性。由於報告內容完整，本研究引用美國化學安全與危害調查委員會BP德州煉油廠調查報告，驗證本研究所提的方法。美國化學安全與危害調查委員會的邏輯樹分析結果，與本研究根本原因分析結果一致。 經由BP德州煉油廠事故的驗證，本研究所提的事故調查方法論可以產出一致性和有意義的結果，因此本研究相信所提的技術可以有效使用於調查與製程相關的意外事故。 Investigating and reporting of unplanned events such as accidents, incidents, and even near-miss events are critical for identifying and controlling potential safety risks as well as fulfilling the requirements of Process Safety Management and OHSAS 18001. The purpose of these activities is not to find fault or lay blame, but rather to identify the basic causes of accidents and incidents so that controls can be put in place to prevent further occurrences. Hence it is essential to look beyond the immediate cause of an accident and look for the contributing factors and basic causes. The most common mistake made by accident investigators is jumping to conclusions on the basis of immediate appearances. A proper investigation has to look into the management system failures. Despite the fact that Article 26 of the Labor Inspection Law has been enforced in Taiwan for about 15 years, the lack of effective accident investigation methodology and the competency level of the safety professionals responsible for accident investigation remain to be one of the key factors hindering the prevention of occupational injuries and process accidents. Hence the primary objective of this study is to provide practical solutions to the problems commonly encountered in incident investigation in Taiwan. A system approach for accident investigation is proposed in this study. The methodology is primarily composed of the steps of collection of evidence and facts, analysis of the information, and root cause analysis. A modified events and causal factors charting, ECFC, is proposed in this study. Modification of the original ECFC makes compiling and organizing accident evidences more effective. It also has the advantage of yielding a set of potential root causes. Because of its detailed description of the affected process units and the background information leading to the catastrophe, the investigation report of the Chemical Safety and Hazard Investigation Board of the US on the BP Texas City refinery accident is used to verify the methodologies proposed in this study. Root cause analysis of the BP Texas City accident is based on the methodology proposed by the Center for Chemical Process Safety of the US. Preliminary analysis of the proposed methodology using the BP Texas City accident yields consistent and meaningful results. It is believed that the proposed technique can be used in investigating process-related accidents.
