Abstract
Temporary speed restrictions (TSR) imposed in railway infrastructure is an important safety measure where circumstances make it necessary to temporarily restrict the speed of trains to less than the normally permitted speed. These circumstances span events such as planned engineer works/maintenance to more extreme measures such as the aftermath of flooding or other extreme weather events that may necessitate a reduction in speed until the initial quality of the infrastructure is re-established. From a traffic standpoint, a reduction in speed would cause longer running time on the affected block (see Figure below).
Figure 1 Schematic breakdown of the speed effects of a TSR
We have studied temporary speed restrictions on The Dovre Line, in Norway. The Norwegian railway consists of 4237 km of track, mainly single tracked (~6% double track). The Dovre Line opened in 1921 and connects the south-east of Norway with mid-Norway, running 485 km (4km double track) between the cities of Eidsvoll and Trondheim. It was electrified in 1970 and has 28 stops with passenger exchange and is under centralised traffic control (CTC). Over the last years, the line has seen significant amount of extreme weather that has caused an extensive number of speed restrictions. We choose to study the speed restrictions imposed in 2012. On the Dovre line in 2012 there were total 54 TSRs of which we analysed 42, which were imposed out on blocks between stations.
We constructed a web-based tool for analyzing the effect of TSRs between stations, for both passenger and freight trains using automatically sampled data from the track circuits in the CTC system. We compared historic traffic data with the period of the TSR. The historic samples have the same duration as the TSR, e.g., if the TSR is active 6 days, the historic samples are also 6 days. The three historic samples are 14 days, 44 days before, and the same period in the year prior to the TSR (Figure 1 shows an example). These three samples deliver consistency in the results and mitigate other effects in the same period.
Figure 2: TSR and historic samples
The boxplots and the T-test shows some screenshots from the system.
Figure 3: Box-plot TSR & historic samples
We have seen that the prioritization of actions between concurrent TSRs and other demands in practice is difficult and simplistic calculations using the theoretical time loss – often present a picture that the rail undertakings not necessarily recognize at the operational end. The effects of TSRs are location dependent and the evaluation of the effect is not a trivial problem. Such a system for easy navigation between before and after effects can allow for more fact-based discussions and evaluations of the traffic effects of speed restrictions. The safety aspect, kept out of this prototype.
The use of empirical data can also be used to validate the acceleration and retardation formulas in relation to the environmental influences, such as uphill or downhill. This could drive an improvement of the theoretical rules applied when empirical investigations are infeasible, but still the primary outcome of this study has been to demonstrate that automatically sampled traffic data can help measure and quantify effects of changes in infrastructure with a higher degree of accuracy than manual reporting and theoretical calculations.
This work has been performed in conjunction with previous work on automatically sampled traffic data – and the use of this in conjunction with the other traffic data-based analytical tools developed in the project displays the benefits of collecting and using such data.
Figure 1 Schematic breakdown of the speed effects of a TSR
We have studied temporary speed restrictions on The Dovre Line, in Norway. The Norwegian railway consists of 4237 km of track, mainly single tracked (~6% double track). The Dovre Line opened in 1921 and connects the south-east of Norway with mid-Norway, running 485 km (4km double track) between the cities of Eidsvoll and Trondheim. It was electrified in 1970 and has 28 stops with passenger exchange and is under centralised traffic control (CTC). Over the last years, the line has seen significant amount of extreme weather that has caused an extensive number of speed restrictions. We choose to study the speed restrictions imposed in 2012. On the Dovre line in 2012 there were total 54 TSRs of which we analysed 42, which were imposed out on blocks between stations.
We constructed a web-based tool for analyzing the effect of TSRs between stations, for both passenger and freight trains using automatically sampled data from the track circuits in the CTC system. We compared historic traffic data with the period of the TSR. The historic samples have the same duration as the TSR, e.g., if the TSR is active 6 days, the historic samples are also 6 days. The three historic samples are 14 days, 44 days before, and the same period in the year prior to the TSR (Figure 1 shows an example). These three samples deliver consistency in the results and mitigate other effects in the same period.
Figure 2: TSR and historic samples
The boxplots and the T-test shows some screenshots from the system.
Figure 3: Box-plot TSR & historic samples
We have seen that the prioritization of actions between concurrent TSRs and other demands in practice is difficult and simplistic calculations using the theoretical time loss – often present a picture that the rail undertakings not necessarily recognize at the operational end. The effects of TSRs are location dependent and the evaluation of the effect is not a trivial problem. Such a system for easy navigation between before and after effects can allow for more fact-based discussions and evaluations of the traffic effects of speed restrictions. The safety aspect, kept out of this prototype.
The use of empirical data can also be used to validate the acceleration and retardation formulas in relation to the environmental influences, such as uphill or downhill. This could drive an improvement of the theoretical rules applied when empirical investigations are infeasible, but still the primary outcome of this study has been to demonstrate that automatically sampled traffic data can help measure and quantify effects of changes in infrastructure with a higher degree of accuracy than manual reporting and theoretical calculations.
This work has been performed in conjunction with previous work on automatically sampled traffic data – and the use of this in conjunction with the other traffic data-based analytical tools developed in the project displays the benefits of collecting and using such data.
