Análisis de confiabilidad en los alimentadores de coque de la planta cogeneradora de ENAP Refinería Bio Bío.
Loading...
Date
2025
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
El sistema de alimentación de coque de la caldera de la planta cogeneradora de ENAP Refinería Bio Bío se soporta en tres alimentadores operando en paralelo. Aunque la demanda de coque puede satisfacerse con solo dos de ellos, esta configuración provoca pérdida de eficiencia térmica y un aumento en las emisiones de dióxido de azufre (SO₂). Entre julio de 2022 y junio de 2025, los tres alimentadores registraron un total de 79 fallas y, en 14 ocasiones, solo uno permaneció operativo, dando lugar a un estado de riesgo operacional. Si llegaran a fallar los tres alimentadores simultáneamente, la planta cogeneradora se detendría y activaría un plan de emergencia que reduce la carga y paraliza parcialmente otras unidades de la refinería. El objetivo general de la memoria de título consiste en identificar las causas raíz de los modos de falla más frecuentes y las brechas en la estrategia de mantenimiento vigente de los alimentadores de coque de la planta cogeneradora de ENAP Refinería Bio Bío, con el propósito de proponer mejoras que reduzcan o prevengan la recurrencia de dichos modos de falla.
Se realizó un Análisis de Modos de Falla y Efectos (FMEA) para desglosar el sistema de alimentación de coque en subsistemas y focalizar el estudio en aquellos subsistemas cuya falla impacta directamente en la función principal de los alimentadores. Después, se caracterizaron los eventos de falla conforme a la norma ISO 14224:2016, integrando datos operativos extraídos de AVEVA PI Vision, registros de SAP y anotaciones de la bitácora de operaciones. Finalmente, se construyó un árbol lógico de fallas (FTA) centrado en los modos de falla que representan más del 80% de las fallas con el objetivo de identificar las causas raíz y sustentar las propuestas de mejora.
El análisis reveló que existe una gran cantidad de fallas tempranas en los tres alimentadores, lo que explica las 14 ocasiones en que solo uno quedó en operación. De los 79 eventos registrados, cinco modos de falla concentraron el 81% de las interrupciones; al profundizar en ellos, se identificaron cuatro mecanismos de daño no cubiertos por la estrategia actual de mantenimiento. En respuesta, se formularon tres propuestas de mejora, cada una con su frecuencia y lista de actividades asociadas. La robustez de la metodología quedó de manifiesto al lograr asignar un modo de falla al 91% de los eventos, y validar el FMEA, los FTA y las propuestas de mejora con personal de mantención.
El análisis de causas raíz de los modos de falla más frecuentes permitió identificar brechas en la estrategia de mantenimiento vigente, validando la hipótesis planteada. Asimismo, se concluyó que el enfoque correctivo aplicado hasta fines de 2024 no fue suficiente para garantizar el correcto funcionamiento de los alimentadores. Las propuestas de mejora, basadas en un plan preventivo focalizado en los mecanismos de daño críticos, ofrecen un esquema sólido para reducir la cantidad de fallas y mitigar las condiciones de riesgo operacional. Como trabajo futuro, se sugiere implementar las propuestas de mejora, monitorear la evolución de los modos de falla y ajustar periódicamente la frecuencia del plan de mantenimiento preventivo en función de los resultados obtenidos. También se recomienda realizar un análisis de los modos de falla que generan mayor indisponibilidad.
The coke feeding system for the boiler at the ENAP Refinería Bio Bío cogeneration plant is supported by three feeders operating in parallel. Although the coke demand can be met using only two of these feeders, this configuration results in reduced thermal efficiency and increased sulfur dioxide (SO₂) emissions. Between July 2022 and June 2025, the three feeders recorded a total of 79 failures, and on 14 occasions, only one feeder remained operational, leading to an operational risk condition. If all three feeders were to fail simultaneously, the cogeneration plant would shut down, triggering an emergency plan involving load reduction and partial shutdown of other refinery units. The primary objective of this thesis is to identify the root causes of the most frequent failure modes and gaps in the current maintenance strategy for the coke feeders at ENAP Refinería Bio Bío’s cogeneration plant, aiming to propose improvements that reduce or prevent the recurrence of these failures. A Failure Mode and Effects Analysis (FMEA) was conducted to break down the coke feeding system into subsystems, focusing the study on those subsystems whose failure directly impacts the feeders' primary function. Subsequently, the failure events were characterized according to ISO 14224:2016 standards, integrating operational data from AVEVA PI Vision, SAP records, and entries from the operation logbook. Finally, a Fault Tree Analysis (FTA) was developed, centered on the failure modes accounting for more than 80% of the recorded events, to identify root causes and substantiate improvement proposals. The analysis revealed a significant occurrence of early failures across all three feeders, explaining the 14 instances where only one feeder remained operational. Of the 79 recorded events, five failure modes accounted for 81% of interruptions; a detailed examination of these modes identified four damage mechanisms not covered by the current maintenance strategy. In response, three improvement proposals were formulated, each including specific maintenance activities and their frequencies. The robustness of the methodology was demonstrated by successfully associating a failure mode with 91% of the events and validating the FMEA, FTA, and improvement proposals with maintenance personnel. The root cause analysis of the most frequent failure modes highlighted gaps in the existing maintenance strategy, validating the hypothesis. It was also concluded that the corrective maintenance approach used until the end of 2024 was insufficient to ensure the reliable operation of the feeders. The proposed improvements, based on a preventive maintenance plan targeting critical damage mechanisms, offer a solid framework to reduce failures and mitigate operational risk conditions. For future work, it is recommended to implement the proposed improvements, monitor the evolution of failure modes, and periodically adjust the frequency of the preventive maintenance plan based on obtained results. Additionally, conducting an analysis of the failure modes causing the greatest unavailability is also recommended.
The coke feeding system for the boiler at the ENAP Refinería Bio Bío cogeneration plant is supported by three feeders operating in parallel. Although the coke demand can be met using only two of these feeders, this configuration results in reduced thermal efficiency and increased sulfur dioxide (SO₂) emissions. Between July 2022 and June 2025, the three feeders recorded a total of 79 failures, and on 14 occasions, only one feeder remained operational, leading to an operational risk condition. If all three feeders were to fail simultaneously, the cogeneration plant would shut down, triggering an emergency plan involving load reduction and partial shutdown of other refinery units. The primary objective of this thesis is to identify the root causes of the most frequent failure modes and gaps in the current maintenance strategy for the coke feeders at ENAP Refinería Bio Bío’s cogeneration plant, aiming to propose improvements that reduce or prevent the recurrence of these failures. A Failure Mode and Effects Analysis (FMEA) was conducted to break down the coke feeding system into subsystems, focusing the study on those subsystems whose failure directly impacts the feeders' primary function. Subsequently, the failure events were characterized according to ISO 14224:2016 standards, integrating operational data from AVEVA PI Vision, SAP records, and entries from the operation logbook. Finally, a Fault Tree Analysis (FTA) was developed, centered on the failure modes accounting for more than 80% of the recorded events, to identify root causes and substantiate improvement proposals. The analysis revealed a significant occurrence of early failures across all three feeders, explaining the 14 instances where only one feeder remained operational. Of the 79 recorded events, five failure modes accounted for 81% of interruptions; a detailed examination of these modes identified four damage mechanisms not covered by the current maintenance strategy. In response, three improvement proposals were formulated, each including specific maintenance activities and their frequencies. The robustness of the methodology was demonstrated by successfully associating a failure mode with 91% of the events and validating the FMEA, FTA, and improvement proposals with maintenance personnel. The root cause analysis of the most frequent failure modes highlighted gaps in the existing maintenance strategy, validating the hypothesis. It was also concluded that the corrective maintenance approach used until the end of 2024 was insufficient to ensure the reliable operation of the feeders. The proposed improvements, based on a preventive maintenance plan targeting critical damage mechanisms, offer a solid framework to reduce failures and mitigate operational risk conditions. For future work, it is recommended to implement the proposed improvements, monitor the evolution of failure modes, and periodically adjust the frequency of the preventive maintenance plan based on obtained results. Additionally, conducting an analysis of the failure modes causing the greatest unavailability is also recommended.
Description
Tesis presentada para optar al título de Ingeniero/a Civil Mecánico/a.
Keywords
Mantenimiento de plantas industriales, Refinación del petróleo, Coque de petróleo