Algerian Journal of Signals and Systems

Yazarlar: Riad BENDIB, Elarkam MECHHOUD, Hanane BENDJAMA, Halima BOULKSIBAT

Konular:-

DOI:10.51485/ajss.v5i1.92

Anahtar Kelimeler:Safety,Probabilistic risk assessment,Outcome frequency,Event tree analysis

Özet: Process plant and chemical processes are complex and large systems, consisting of thousands of devices interacting with each other [76][45]. The structure of this type of industrial plants lead to difficulties in process control, hence deviations from the operation objectives or desired states. These deviations creates abnormal situations that drive products out of specification, increased operational costs, shut downs and even worse they can cause accidents, which may lead to damages of equipment, environment and affect the health. identifying hazards is fundamental for ensuring the safe design and operation in process plants. Based on the idea of learning from accidents, several techniques and standards are available to identify hazardous situations and help companies to build up strategies to avoid the hazards. In this s work we present a study based on two methods BOWTIE and HAZOP methods applied for an LPG a plant for gas separation located in SKIKDA refinery which is considered as the most important refinery in Algeria where the treatment capacity reaches more than 15 millions tones per year of crude oil. Several recommendations raised from our study to improve the safety of the plant particularly since it is considered as an old plant start working since 1980. The study is completed by simulating the deduced dangerous scenarios using ALOHA software.

---