Currently many railway companies operate on a mixed infrastructure, i.e. trains with big speed differences (e.g. freight and high speed passenger trains) run on the same tracks. This has a number of drawbacks: low capacity, additional infrastructure costs for passing lanes etc.

Therefore infrastructure operators try to “demix” their infrastructure by restricting trains to different tracks according to average speed. Figure 1 shows (in a very simplified manner) the principle of line demixing and the consequent increase in line capacity.

This can be a way to increase system capacity without investing in new infrastructure, thus lowering specific infrastructure costs.

Presently many railways recur to temporal demixing, i.e. freight trains running during night hours, passenger trains during day-time. In many cases, the spatial demixing described here is however a much better and much more customer-friendly strategy.

Figure 1: Principle of line demixing and capacity effect (simplified)

Source: IZT

A positive effect of demixing on the energy efficiency of train operation is to be expected since demixing improves traffic fluidity by avoiding train stops or decelerations caused by speed differences.

There are however different effects brought about by demixing that have to be clearly distinguished:

1. Scheduled stops on the way due to slower trains letting pass faster trains are clearly reduced in a demixed infrastructure.
2. Unscheduled stops due to delayed trains impeding other trains may be reduced by demixing (cf. Figure 2).
3. Increased train density in a demixed infrastructure with raised capacity may increase number of train conflicts due to delays.

Figure 2: Schematic visualization of effect of train delys in mixed vs. demixed infrastructure (simplified)

Source: IZT

Effects 1 and 2 reduce the probability of train stops outside stations in a demixed infrastructure (as compared to the mixed case), effect 3 points in the opposite direction. It is very difficult to make a general statement on the dominance of one effect over the other and the total net effect.

Experts however tend to assume that effects 2 and 3 compensate each other yielding a zero net result. Effect 1 could then slightly improve overall “stopping statistics”.

Conclusion

Energy efficiency effects through demixing depend strongly on the individual network situation and trains densities. In most cases, the net effect will rather be a small positive one. It will however be extremely difficult to give reliable numbers in this context.