A HIRARC model for safety and risk evaluation at a hydroelectric power generation plant

Abstract

There are many formal techniques for the systematic analysis of occupational safety and health in general, and risk analysis in particular, for power generation plants at hydroelectric power stations. This study was initiated in order to create a HIRARC model for the evaluation of environmental safety and health at a hydroelectric power generation plant at Cameron Highlands in Pahang, Malaysia. The HIRARC model was used to identify the primary and secondary hazards which may be inherent in the system which were determined as a serious threat for plant operation and maintenance. The primary tools of the model consisted of, generic check-lists, work place inspection schemes which included task observation and interview, safety analysis as well as accident and incident investigation. For risk assessment, the Likert scale was complemented by the severity matrix analysis in order to determine the probability and extent of safety and health at the study power generation plant. These were used to identify and recommend control measures which included engineering and administrative aspects as well as the use of personal protective equipment (PPE). A total of forty-one important hazard items were identified in the system at target power generation plant. These hazards were mainly identified by means of checklists which were sourced from literature and subsequently customized for the current purpose. Risk assessment was conducted by initially classifying the hazards into three levels such as Low, Medium and High. Generally 66% of the hazards identified were at low risk, 32% at medium and 2% at high risk. This indicated that there was sufficient awareness and commitment to safety and health at the study power station. Meanwhile the Power Station was also certified by MS 1722:2005, OHSAS 18001, MS ISO 14001:2004, MS ISO 9001:2000 and scheduled waste regulation 2005 which give credibility to the current study in creating a working model which may find widespread application in the future.