| Abstract: | Flare gas utilization in a cogeneration plant is an attractive proposition considering its environmental and economic incentives. Evaluation of the operational risk of integrating flare gas with cogeneration is complex due to the uncertainty in flare gas quality and process conditions. The current study delves into the change in operational risk after modifying the existing cogeneration process with the addition of fuel from flare gas. Based on the process hazards evaluation, the current study identified two critical loss control events (or top events) - boiler gas temperature exceeding operating design temperature and rich fuel mixture in the boiler firebox. The underlying causes that may contribute and lead to these loss control events were identified using fault trees and were updated to the existing cogeneration scenarios. Similarly, different consequential events that may arise from the loss control events were analysed using event trees with existing system safeguards. A Bayesian network model with its explanatory power mapped all the identified dangerous scenarios from the fault trees and event trees to predict integrated systems reliability and diagnose causal factors. Bayesian Network analysis illustrates the dynamic cause-effect relationship and determines the risk escalation due to the changes in the composition of flare gas that is fed to the boiler. The presence of a higher percentage of hydrogen (above 40 mol%) in the flare gas escalates the risk of lean air to fuel ratio in the boiler firebox and increases boiler radiation zone duty. These conditions are detrimental to the boiler firebox operation and can result in critical scenarios such as flame impingement and tube rupture. Additionally, other consequences-a steam explosion and boiler stack explosion were also investigated. However, their probability of occurrence was relatively insignificant with the given frequency of flare gas utilization in the cogeneration system. |
