Functional dependency between the number of wagons derailed due to wagon or track defects and the traffic factors

Aim. Rolling stock derailment is one of the most hazardous transportation incidents. Depending on the gravity of the consequences they may also be classified as accidents or train wrecks. The consequences of a derailment may vary from routine maintenance of the track and one or two wagons to an overhaul of the track and depot repairs of three or more wagons, as well as loss of cargo and long interruption of service. It must be noted that beside the damage to infrastructure and rolling stock caused by derailments there is a risk of environmental disaster. The Russian Federation along with the US, China and India has some of the world’s longest rail networks that in places border with environmentally sensitive areas, e.g. national reserves and parks. Therefore, if a train carries hazardous cargo, e.g. gasoline or toxic gases and some of its wagons derailed, the harm related to the repair or decommissioning of rolling stock, track and possible loss of cargo may be aggravated by the damage caused by an environmental disaster that would cause great material losses to JSC RZD. In this context it appears to be of relevance to evaluate the functional dependency between the potential number of cars derailed and various factors, e.g. speed or amount of cargo carried by the train, for subsequent preparation of recommendations for the reduction of the potential number of derailed cars and, subsequently, reduction of possible harm. Methods. Probability theory and mathematical statistics methods were used, i.e. maximum likelihood method, negative binomial regression. Results. For various groups of incidents, i.e. derailment as the result of wagon or locomotive unit malfunction out of switch, derailment as the result of rail malfunction out of switch, derailment at a switch not caused by previous derailment, specific functions of the average number of derailed wagons are identified. The paper shows a formula that allows – under a defined set of various factors, e.g. train speed, plan and profile of track, length and mass of the train – identifying the distribution series of the number of derailed wagons. Conclusions. The preliminary analysis of available Russian freight train derailment records it was shown that for various groups of transportation incidents the descriptive statistics of respective samples significantly differ, which is also the case for the US records. The construction of a functional dependence between the average number of derailed wagons and various traffic factors due to malfunction of wagons or locomotive units out of switches, it was identified that the available records do not suffice to forecast the number of derailed wagons in tangents. Mathematical models with a low superdispersion parameter were constructed for derailments due to track malfunction out of switches and derailments at switches. 

1. <http://safetydata.fra.dot.gov/>

2. <http://www.indianrailways.gov.in/>

3. Goriainov AV, Zamyshliaev AM, Platonov EN. Analysis of effects of factors on the damage caused by accidents in transportation using regression models. Dependability 2013;2:126-144.

4. Liu X, Saat MR, Qin X, Barkan CPL. Analysis of U.S. freight-train derailment severity using zero-truncated negative binomial regression and quantile regression. Accident Analysis and Prevention 2013;59:87-93.

5. Liu X, RapikSaat M, Barkan CPL. Freight-train derailment rates for railroad safety and risk analysis. Accident Analysis and Prevention 2017;98:1-9.

6. Anderson RT, Barkan CPL. Derailment probability analysis and modeling of mainline freight trains. In: Proceedings of the 8-th International Heavy Haul Conference, Rio de Janeiro (Brazil): International Heavy Haul Association; 2005. p. 491–497.

7. Bagheri M, Saccomanno F, Chenouri S, Fu LP. Reducing the threat ofin-transit derailments involving dangerous goods through effective placement along the train consist. Accident Analysis and Prevention 2011;43:613–620.

8. Saccomanno FF, El-Hage S. Minimizing derailments of railcars carrying dangerous commodities through effective marshaling strategies. Transportation Research Record 1989;1245:34–51.

9. Koenker R, Hallock KF. Quantile Regression. Journal of Economic Perspectives 2001; 15(4):143-156.

10. DeGroot MH, Schervish MJ. Probability and Statistics. 4th ed. Addison-Wesley; 2012.

11. Cameron AC, Trivedi PK. Essentials of Count Data Regression in A Companion to Theoretical Econometrics. Baltagi BH, editor. Blackwell Publishing Ltd; 2003.

12. Zamyshliaev AM, Ignatov AN et al. Ob otsenkekolichestvavagonov v skhodepripoezdnoyrabotenaosnovefaktornykhmodeley [On the evaluation of the number of derailed wagons in operation based on factor models]. In: Proceedings of the Sixth Science and Engineering Conference Intelligent Control Systems in Railway Transportation. Computer and mathematical modeling. ISUZHT 2017. 2017. p. 132-135.