Thesis Desarrollo y validación de una plataforma de simulación para diagnosticar la capacidad efectiva del sistema de atención de camiones del Terminal Marítimo Oxiquim Quintero
No Thumbnail Available
Date
2026-10-08
Authors
Journal Title
Journal ISSN
Volume Title
Program
Ingeniería Civil Industrial
Departament
Campus
Campus Casa Central Valparaíso
Abstract
El Terminal Marítimo Oxiquim Quintero enfrenta un problema recurrente de congestión logística en su sistema de atención de camiones, que se manifiesta en tiempos de espera elevados, saturación de espacios operativos y riesgos de seguridad asociados al tránsito peatonal en zonas operativas, sin que exista, al inicio de este trabajo, una determinación sistemática de su capacidad efectiva. El propósito de este trabajo es desarrollar y validar una plataforma de simulación del sistema de atención de camiones del terminal, y aplicarla al diagnóstico de la capacidad efectiva de sus principales líneas de producto y a la evaluación de escenarios de demanda y de mejora operacional, de modo que sirva de línea base para un futuro modelo integral. Para ello se combinaron cuatro componentes: un análisis estadístico de aproximadamente 111.000 registros operacionales históricos de la empresa (incluyendo un modelo predictivo del tiempo de servicio, verificación de la Ley de Little, y un modelo analítico de teoría de colas); un simulador visual de eventos discretos desarrollado en Python; un motor de simulación estadístico en R, validado progresivamente contra los datos reales; y una herramienta interactiva de apoyo a decisiones. El análisis estadístico permitió establecer que un modelo analítico de colas simple subestima la espera real en dos órdenes de magnitud, brecha explicada por una restricción de agrupamiento (pooling) entre mesas de carga y por la no estacionariedad de la demanda. El motor de simulación desarrollado, que sí incorpora ambos mecanismos, logra capturar entre el 61% y el 84% de la espera real observada según la línea de producto. Se determinó además la capacidad efectiva de las principales líneas de producto: ácido resulta ser la línea más ajustada en su hora de mayor demanda (24,1% de margen), con químico y GLP como casos secundarios; químico y NASH (sulfhidrato de sodio) aceleran su espera de forma no lineal a partir de un crecimiento de demanda de entre 150% y 300%, mientras que la isla de hidrocarburos requiere entre 12,5 y 15 veces la demanda actual para un comportamiento equivalente. Finalmente, (...).
The Oxiquim Quintero Marine Terminal (TMQ) faces a recurring logistics congestion problem in its truck service system, manifested in long waiting times, saturation of operational spaces, and safety risks associated with pedestrian traffic in operational areas, with no systematic determination of its effective capacity available at the outset of this work. The purpose of this thesis is to develop and validate a simulation platform for the terminal's truck service system, and to apply it to the diagnosis of the effective capacity of its main product lines and to the evaluation of demand and operational improvement scenarios, so that it may serve as a baseline for a future integrated model. To this end, four components were combined: a statistical analysis of approximately 111,000 historical operational records from the company (including a predictive model of service time, verification of Little's Law, and an analytical queueing model); a visual discrete-event simulator (DES) developed in Python; a statistical simulation engine in R, progressively validated against real data; and an interactive decision-support tool. The statistical analysis established that a simple analytical queueing model underestimates real waiting time by two orders of magnitude, a gap explained by a pooling restriction across loading bays and by the non-stationarity of demand. The simulation engine developed, which incorporates both mechanisms, captures between 61% and 84% of the observed real waiting time depending on the product line. The effective capacity of the main product lines was also determined: sulfuric acid is the tightest line at its peak hour (24.1% margin), with chemicals and LPG as secondary cases; chemicals and sodium hydrosulfide (NASH) exhibit non-linear acceleration of waiting time from a demand growth of between 150% and 300%, whereas the hydrocarbons island requires between 12.5 and 15 times its current demand for equivalent behavior. Finally, (...).
The Oxiquim Quintero Marine Terminal (TMQ) faces a recurring logistics congestion problem in its truck service system, manifested in long waiting times, saturation of operational spaces, and safety risks associated with pedestrian traffic in operational areas, with no systematic determination of its effective capacity available at the outset of this work. The purpose of this thesis is to develop and validate a simulation platform for the terminal's truck service system, and to apply it to the diagnosis of the effective capacity of its main product lines and to the evaluation of demand and operational improvement scenarios, so that it may serve as a baseline for a future integrated model. To this end, four components were combined: a statistical analysis of approximately 111,000 historical operational records from the company (including a predictive model of service time, verification of Little's Law, and an analytical queueing model); a visual discrete-event simulator (DES) developed in Python; a statistical simulation engine in R, progressively validated against real data; and an interactive decision-support tool. The statistical analysis established that a simple analytical queueing model underestimates real waiting time by two orders of magnitude, a gap explained by a pooling restriction across loading bays and by the non-stationarity of demand. The simulation engine developed, which incorporates both mechanisms, captures between 61% and 84% of the observed real waiting time depending on the product line. The effective capacity of the main product lines was also determined: sulfuric acid is the tightest line at its peak hour (24.1% margin), with chemicals and LPG as secondary cases; chemicals and sodium hydrosulfide (NASH) exhibit non-linear acceleration of waiting time from a demand growth of between 150% and 300%, whereas the hydrocarbons island requires between 12.5 and 15 times its current demand for equivalent behavior. Finally, (...).
Description
Keywords
Terminal marítimo, Simulación de eventos discretos, Teoría de colas, Capacidad operacional, Congestión logística, Análisis estadístico, Apoyo a la toma de decisiones
