Predicting and measuring service disruption recovery time in railway gravity hump classification yards

Abstract

Planned maintenance and unplanned incidents cause service disruptions in freight railway classification yards, creating congestion, delaying railcars, and even impacting mainline operations. Understanding the recovery time and lingering performance impacts of yard disruptions is vital for the industry to plan disruption responses, promote efficient resource utilization, and improve resiliency. This paper compares two major types of yard disruptions (temporary closures of hump process and pulldown process) and quantifies the recovery pattern, measured by multiple performance metrics. The authors propose an analytical approach for estimating classification yard recovery time as a function of disruption duration and baseline capacity utilization. To validate the hypothetical approach, a series of experiments are conducted across a wide range of disruption durations and throughput volumes in a representative hump classification yard simulation model constructed using AnyLogic. The results indicate that recovery time is proportional to shutdown duration with a near constant recovery rate, and recovery rate increases approximately exponentially with throughput volume. These results are consistent with the hypothesized analytical relationships, suggesting that yard capacity may be estimated from disruption recovery rate. The methodology developed also enables future studies on interactions between yards and mainlines and developing planning-level parametric models of classification yard capacity and performance.