Hazard Analysis and Preventive Measures for the PTA Manufacturing Process

• Main Authors: Liao, Hung-Pen, 廖本弘
• Other Authors: Kam,Tai-Yan
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/5x786z

碩士 === 國立交通大學 === 工學院產業安全與防災學程 === 107 === In this study, the safety issue of the Purified Terephthalic Acid (PTA) manufacturing process of a petrochemical plant in Taiwan is studied. In producing PTA, the petrochemical plant uses as well as produces high-pressure gases during the production and has thus been categorized as Class C dangerous work place. This study applies the Preliminary Hazard Analysis (PHA), the Hazard and operability study (HazOp), and the Layer of Protection Analysis (LOPA) to evaluate the process hazard and performs the Fault Tree Analysis (FTA) to evaluate the main causes of reactor abnormality. The high-pressure gas in the production area is vulnerable to explosion risks. Through HazOp Analysis, a total of 55 items in the production area including 25 preventive measures have been analyzed. The results show five items have been ranked Level 5 severity. LOPA is further applied to calculate the failure probability of each independent protection layer in order to effectively control the risks after the improvement. With the application of the new measures, the risks have been reduced to an acceptable level. FTA is applied to analyze the causes of reactor explosion and finds the probabilities of reactor explosion and fire are 8.8310-20 and 6.8910-32, respectively. Finding the main reasons for reactor explosion and fire, the study establishes plans in different stages to reduce the error rate. In the short-term plan, patrolling system is reinforced and digitalized; the mid-term plan corrects the maintenance and annual repair plans; the long-term plan is to purchase spare parts for those with high failure rates. By doing so, the failure causes can be analyzed effectively and the maintenance cost can be reduced. Based on the failure rates, the changing times of parts can be evaluated and the purchase of spare parts can effectively reduce inventory and costs. Patrolling can improve and predicts risks directly. The smart patrolling system adopts the Near Field Communication (NFC) mobile technology program to replace manual meter reader. A total of 300 paper forms in the factory are converted into 60 electronic forms. The smart patrolling system is installed on the personal digital assistant (PDA) of meter readers. With the implementation of smart patrolling, the abnormality severity of 3 production areas has been reduced to the acceptable level. The results obtained in this study have the potential to serve as a reference for other domestic petrochemical or chemical plants to re-evaluate their risk assessment tools and examine process risk as well as the internal risk reduction measures of the enterprise.