Risk management for safety operation utilizing virtual reality simulation supported by intelligent HAZOP analysis

Mad Sahar, Nan (2010) Risk management for safety operation utilizing virtual reality simulation supported by intelligent HAZOP analysis. Doctoral thesis, Okayama University.


Download (11MB) | Preview


Ensuring safe operability and minimizing risk is the key component to prevent negative impact in all industries dealing with toxic, reactive, flammable and explosive materials. HAZOP (Hazard and Operability), a preliminary and systematic approach for identifying hazards has been unquestionably successful in reducing incident of hazards by mitigating the consequence of major accident in the industrial process facilities. However, laborious work, time and cost are the shortcoming in performing and maintaining HAZOP analysis. Many research works on HAZOP automation are available, yet the traditional approach is still widely used by plant operators. The traditional method only covers parts and aspects of a specific plant type rather than generalizing to fit many plant types. In HAZOP analysis of chemical process industries (CPI), process analysis can be divided into two groups - defined or routine process, which roughly occupies 60-80% and predefined or non routine process, which occupies 20-60% of HAZOP analysis. Thus leading towards the significance of having safety information as update and accessible as possible. In recent years, computer hardware capable of developing and running virtual reality model has become more affordable for middle and small scale CPI. Consequently, virtual reality has been proposed as a technological breakthrough that holds the power to facilitate analysis. The ability to visualize complex and dynamic systems involving personnel, equipment and layouts during any real operation is a potential advantage of such an approach. With virtual reality supporting HAZOP, analysis which often solely relied on expert imaginative thinking in simulating hazard conditions, will aid understanding, memory retention and create a more interactive analysis experience. In focusing assessment for safety operator and safety decision maker, we present a web-based HAZOP analysis management system (HMS) to help HAZOP team and related individuals to perform revision, tracking and even complete HAZOP analysis without management bureaucracy. Besides, depending solely on expert imaginative thinking of scenario using P&ID, this work will develop a dynamic visual model which brings to the user a different view of consequent and subsequent to an accident and will further enable three dimensional analyses of effects. This approach will prevent ‘miss looks’ due to ‘paper-based’ view. We also present Virtual HAZOP Training system, a risk-managing virtual training concept supported by intelligent HAZOP proposed to eliminate analysis redundancies and bring static ‘paper-based’ analysis to more dynamic and interactive virtual analysis simulation. However, the efficiency of VR simulator depends on the scenario accuracy to the real world that can be simulated. We introduce the system’s artificial intelligent engine responsible for retrieving the most accurate and highest possibility ‘to-happen’ scenario case. A fuzzy – CBR method enables the engine to classify and use real past scenarios combined with suitable parameters in creating a defined scenario. This method resolves issues in balancing between computational complexity and knowledge elicitation Reactor section in a vacuum gas oil hydrodesulphurization (VGO HDS) process is used as the case study to illustrate the performance of the proposed system. The wide usages of HDS unit in the petroleum refining industry play important roles in chemical plant incidents happening worldwide. HAZOP analysis management system in average manages to reduce more than half the time required in performing HAZOP analysis compares to traditional method. With the proposed system, operator is able to optimally use safety information in HMS to prevent common and repetitive mistakes. Virtual process and accident simulator available in virtual HAZOP training system help to improve safety operator estimate overall impact towards equipment, operator and environment during process 20-35% better. This system is expected to be the main foundation for Virtual Reality simulator research in analyzing accident caused by human factor. Asides providing better and healthier working environment, negative profitability impact which influence not only the company that runs it but also the world economy due to byproduct shortage, can be avoided.

Item Type: Thesis (Doctoral)
Subjects: T Technology > TP Chemical technology
Depositing User: Mrs. Sabarina Che Mat
Date Deposited: 03 Feb 2022 01:58
Last Modified: 03 Feb 2022 01:58
URI: http://eprints.uthm.edu.my/id/eprint/3671

Actions (login required)

View Item View Item