When evaluating the energy efficiency of running MUs vs. locomotive-hauled stock, the following issues have to be addressed:

• The weight per seat in MUs is smaller than in locomotive-hauled trains.
• EMUs offer better performance in regenerative braking than loco-hauled stock
• When taking load-factors into account (i.e. going from a seat-specific to a passenger-specific perspective), the outcome of the comparison is controversial.

Weight per seat

A comparison between the ICE 2 and the ICE 3 indicates the effect of decentralised traction on mass per seat in high-speed operation. This is illustrated in the following table:

	ICE 2	ICE 3
Mass	410 tons	409 tons
Length	205 m	200 m
Seats	391	415
Mass per train length	2000 kg/m	2045 kg/m
Mass per seat	1049 kg/seat (100%)	986 kg/seat (94%)

Source: Rahn 2001, IZT calculations

This comparison shows that in ICE 3 weight per seat is reduced by 6% compared to the ICE 2 having centralised traction. This is only due to the increased seating capacity since the mass per train length is not reduced. In shorter trains, the difference in weight per seat between MU and loco-hauled solutions will be even more pronounced. In general values, mass reduction between 5 and 10% will be realistic. Taking elasticities into account, this yields energy savings of 1 to 5 %, depending on the type operation.

Regenerative braking

Generally, EMUs have a better regenerative braking performance than loco-hauled trains, since more axles are powered. The higher the motor power and the more axles are powered, the more energy may be recovered. This effect is difficult to quantify but may be 5% or more in many cases.

Passenger-specific perspective

While some experts state that loco-hauled stock offers more flexibility to adapt train length to varying demands, others claim that short MUs which can be coupled to form longer units are unbeaten as far as flexibility is concerned. It is obviously true that loco-hauled trains offer the highest flexibility, since any number of cars is feasible. In practice however, barriers such as coupling efforts and complicated vehicle logistics often impede flexibility of train formation. Therefore the flexibility advantage of loco-hauled stock seems to be more of a theoretical nature.

Conclusion

There is a clear energy advantage of MUs over loco-hauled stock being in the order of 5 - 10%. As far as flexibility isues are concerned a comparison between MU and loco-hauled is much more difficult to establish, but experience from several operators shows that short MU stock with automatic coupling shows the best flexibility in everyday operation.