An Investigation into the Highway-Rail At-Grade Intersection Safety—Case Study: Zambia Railways Rail Network

The transportation sector in Zambia faces many challenges, one of them being the deteriorating safety situation at highway-rail at-grade intersections (HRAGIs), sometimes called level crossings. Therefore, the aim of this research was to carry out an investigation into HRAGIs safety on the Zambia Railways rail network and find ways for improvement. Three critical (black spots) Class-A level crossings in the northern region, i.e., Bwana Mkubwa in Ndola, Lukanda in Kapiri Mposhi and Kabwe Main in Kabwe were selected for investigation. The methodology involved field surveys, visual inspection, accident and hazard prediction, interviews and driver behavior assessment. The adopted Florida Accident and Safety Index Model, revealed that the safety indices for all the three level crossings, without and with gates, were unsafe and below the recommended marginal threshold of 60%. The safety indices (without gates) were, for Bwana Mkubwa (51.49%), Lukanda (48.34%) and Kabwe Main (53.31%) and were (with gates) 53.01%, 50.02% and 54.73% for Bwana Mkubwa, Lukanda and Kabwe Main respectively. On the other hand, the USDOT Accident Prediction Model initial accident prediction showed low accident probabilities, compared to observed accident rates, with an average of 0.20 accidents per year for the three HRAGIs with the second accident prediction showing Lukanda as more vulnerable at 0.28, compared to 0.21 for Bwana Mkubwa and 0.15 for Kabwe Main predicted accidents per year. Nevertheless, the USDOT predicated results were consistent with the observations made in the USA. The low predicated values could be attributed to low daily train volumes, reduced maximum legal train speeds of 40 km/h and enforced lower speeds (15 - 20 km/h). Other factors such as human error, vandalism of infrastructure and also lack of lighting at all the three crossing could have played a major role. Consistent data collection is needed to realize the development of local models through statistical modelling. Also, continuous maintenance and upgrades of HRGAIs from passive to active status including driver behavior sensitization is recommended going forward.