In the freight sector, the specific advantages of rail are ideally exploited by long trains transporting heavy low-value mass goods from point A to point B. The specific advantages of road transport are best realised by small high-value goods that have to be transported in small quantities.

During the last decades, the latter type of freight has constantly increased while mass goods have lost importance. For smaller amounts of cargo the conventional production system in railways has been the one illustrated in solution 1 of Figure 1. This system is cost and time-consuming since the individual units have to be coupled and decoupled in shunting points and often have to wait until enough units have gathered in order to form a long train on the main relation. These problems are one of the reasons why the modal split has changed in favour of road transport.

Figure 1: Customer needs and how they are met by different production systems

Source: IZT (based on Frederich, Lege 1996)

The most obvious solution to this problem is to make freight trains more truck-like, i.e. replace long loco-hauled trains by smaller units with a high degree of modularity and flexibility (due to rapid automatic coupling and decoupling etc.).

These shorter units can be realised in different ways:

• Short conventional loco-hauled freight trains
• CargoSprinter consisting of multiple platforms, the end ones of each group are powered by a small diesel motor. The intermediate platforms are unpowered. Several of these trains can be linked together and run in MU (multiple unit) configuration.
• Individual self propelled freight cars: Each wagon is powered and runs independently (usually requiring driverless operation).

The following refers to self-propelled freight cars since they are the most radical realisation of the concept of modular and flexible freight trains.

Self-propelled freight cars have a propulsion unit on-board. This can be realised by electric or diesel traction, the latter being cheaper and better suited to freight traffic running frequently on non-electrified tracks. Self-propelled freight cars could be used for a direct point-to-point delivery of small freight quantities. From an economic point of view, this can economically only be realised by driverless operation. This requires

• autonomous navigation
• automatic avoidance of conflict with other vehicles (replace security control procedure based on fixed railroad sections and visual signalling by telematics applications)
• new shunting methods due to application of telematics and self-powered waggons

   

• A cost-effective design and layout of self-propelled freight cars is a future challenge for R&D.
• For the driverless operation of self-propelled freight units there exists no mature system yet. This problem seems principally solvable by future R&D.

Competition

•  Meet the requirements of changing freight markets and win back market share from road transport.
• Make present production system in rail freight transport more cost-effective.

Cost efficiency

No drivers, point-to-point transport, no cost-intensive cargo transport centres freight handling.

Time efficiency

Point-to-point transport, no time-intensive coupling and reloading processes.

Customer service

In the future the customer himself could send self-propelled cars on their way whenever and whereever desired. The goal is to make sending freight for the customer as easy as mailing a letter.

Transition costs and system inertia

The introduction of self-propelled and self-organising freight cars would represent a system change in rail-bound freight transport. Not only high transition costs but also uncertainties about the new system constitute high hurdles.

Initial costs

The load-specific initial investment would presumably be higher for powered freight cars than for the loco-hauled production system.

Safety

A fleet of self-organising driverless freight cars operating on a mixed infrastructure poses a number of serious safety problems that are currently being studied.

Study by Rauschenberg

There is a theoretical study (Rauschenberg (no year)) examining the external effects of the new system. A very rough assumption is made about running resistance of self-propelled freight cars:

• rolling resistance is assumed to be equal to conventional freight trains
• air resistance is assumed to be equal to road trucks

These assumptions, which seem reasonable for a first approach, lead to the running resistance shown in Figure 2. For an average speed of 80 km/h, this yields the comparison given in Table 1. It shows that the running resistance of self-propelled freight cars is roughly 3 times that of a freight car in a loco hauled train

Figure 2: Comparison of running resistance of conventional rail transportation, self-propelled rail vehicles and road transportation

Source: Rauschenberg 2000

Table 1: Comparison of running resistance at 80 km/h for conventional rail transportation, self-propelled rail vehicles and road transportation (figures give load-specific resistance force, i.e. resistance force per unit of weight force of load)

	Road truck	conventional freight    train	Self-propelled freight    car
Rolling resistance	1,037 %	0,4306 %	0,4306 %
Air resistance	1,029 %	0,09574 %	1,029 %
Running resistance	2,066 %	0,5263 %	1,459 %

Source: IZT, data from Rauschenberg 2000

Estimate using data from Vollmer 1989

Since a self-propelled freight car will aerodynamically behave similar to the first car in a train the data from Vollmer 1989 can be used to make an estimate on the additional air resistance met by self-propelled freight cars. Since the first car in a train configuration roughly causes 4 - 10 times as much air drag as the following ones, this is also a good estimate for the difference between self-propelled freight cars and loco-hauled stock. This is in reasonable accordance with the factor 10 assumed by Rauschenberg (cf. second line in Table 1)

Conclusion

Assuming that the (load-)specific air drag of self-propelled cars is about 4 - 10 times higher than that of loco-hauled stock, and taking into consideration that air drag accounts for about 50 % of total energy consumption in freight transport, one derives a total energy consumption increased by a factor of 2 to 5.

The additional energy demand will be reduced due to the fact that self-propelled cars travel shorter distances owing to point-to-point service and do not need additional freight handling along the route.

Energy efficiency of self-propelled could be improved by two factors:

• Running resistance can be substantially reduced by automatic coupling of self-propelled units to form long trains on major routes. This way self-propelled freight cars could exploit the aerodynamic advantages of long freight trains without sharing their drawbacks.
• The same holds for the concept of Virtually coupled trains with much higher requirements for train formation.
• If self-propelled freight cars are equipped with electric traction, they can use regenerative braking, which would cut the effect of mass acceleration. This is of course a cost issue, since electric propulsion will be more expensive than diesel propulsion.

The environmental assessment of self-propelled freight cars requires a new perspective. It is hardly reasonable to compare energy efficiency of the new train concept with conventional freight trains since self-propelled freight cars compete for a market segment presently dominated by road trucks rather than rail transport.

Therefore the concept of self-propelled freight cars can be seen as an energy-efficient option even though the comparison with conventional freight trains does not favor them.

Assuming that the system of self-propelled freight cars can be realised which uses automatic train formation on main routes, optimised point-to-point logistics and load management, the overall energy consumption could be comparable or even beat conventional railway freight systems.