Publications

Poorly managed alarms and improper design of alarm system have contribute to major industrial incident. Some examples are fire explosion on Milford Haven, 24th July 1994 where it flooded by alarms at the rate of one every two or three seconds resulting operator cancelling those alarms due to the nuisance (Health and Safety Executive, 1997), other case was from Texas City Refinery explosion, 23rd March 2005 where the high level alarm was ignore while the secondary high alarm was faulty (BP Executive Summary, 2005), and recent incident of Fatal Gas Well Blowout in Pryor Trust gas well Pittsburgh Country 22nd January 2018 where the entire alarm system was disabled by rig personnel due to nuisance condition (U.S. Chemical Safety and Hazard Investigation Board,2019). Banyu Urip facility have put significant effort on improving alarm effectiveness. This paper describes Banyu Urip alarm management workflow and improvement consisting of data collection, alarm analysis (identify bad actor), short term repair/improvement, alarm review & rationalization, management of change process, re-evaluation of operating envelopes, and high focus for implementation and stewardship. Alarm management tools and dashboard significantly helps the process. There are 4 KPIs that periodically reviewed: number of alarm rate, standing alarm, shelved alarm and critical operating parameter. This paper also provide Banyu Urip plant alarm management journey that started by first cold eyes review found imbalance of 3 panel operator load. Reconfigure asset assignment significantly improve the alarm rate as well as plant stability. The journey continue by multiple approach depend on the case of alarm such as nuisance alarm and standing alarm consisting alarm rationalization, alarm suppression logic, equipment out Of service logic, time delay implementation, control loop tuning, and deadband setting. Other than that, manual shelving process also utilized for nuisance alarms short term action caused by equipment or sensor problem, as long as mitigation exists, communicated on every shift handover and directly followed up to maintenance order list. There are also some cases where known equipment issues waiting for long term resolution causing continuous alarms. Risk screening process help all parties aware on incremental risk during the interim period. Having above alarm management improvement workflow, CPF site team have able to reduce down the alarm rate to stable criteria (<12 / hours) and standing alarm continue to decrease to 10/console. Finally, this paper also describes how alarm management practices can be also utilized as supplementary surveillance dashboard to improve plant reliability