Abstract
M. Ing.
This dissertation proposes a methodology for the analysis of the operational risks that
are caused by railway infrastructure component failure. The objective of this
methodology is to assist the engineers that manage railway infrastructure maintenance
to forecast the following:
• the frequency of operational risk events that are caused by railway
infrastructure failure;
• the cost of rehabilitating railway infrastructure after an operational risk event
that was caused by railway infrastructure failure; and
• the impact that railway infrastructure maintenance strategies have on the
frequency and cost of operational risk events that are caused by railway
infrastructure failure.
A brief literature study of operational risk analysis is presented.
The proposed operational risk analysis methodology involves the identification of the
operational risks that are caused by railway infrastructure failure and causal modelling
using Bayesian network causal models.
The proposed operational risk methodology is applied in a case study concerning a
railway company (called African Railways Ltd as a pseudonym for the sake of
confidentiality). The train derailments that are caused by infrastructure component
failure in a particular region are analysed in the case study. The case study presents
historical data and the results of a questionnaire that was used during face-to-face
individual interviews with three track maintenance experts. The frequency of train
derailments, the cost of rehabilitating railway infrastructure after train derailments and
the impact of railway infrastructure maintenance on these two issues are forecasted.
The case study concludes with a comparison of the forecasted and actual frequency of
train derailments and cost of rehabilitating the railway infrastructure after a train
derailment.