IEEE Electrification Magazine - September 2014 - 34

Flat-Out Power and Speed Profiles

Power (kW)

2,000
1,000

10 kWh

0
23.7 kWh

-1,000

4 kWh

-2,000
0

50
60
Time (s)
(a)

100

70
Speed (km/h)

60
50
40
30

Acceleration

Cruising

Braking

10
0

50
60
Time (s)
(b)

100

Figure 1. The consumption and regeneration in a metropolitan train running between two passenger stations.

voltage in substations to prevent possible voltage dips.
In addition to avoiding low-voltage levels, this practice
reduces transmission losses. However, it also reduces
the voltage range that is available for regeneration, and
the line receptivity deteriorates. Recent publications
propose the use of substation-transformer taps to
change the no-load voltage a few times per day,
improving the total efficiency of the system.
Despite these problems, there is an important potential
for energy-efficiency improvement in dc-electrified railway
systems, which are very common and fundamental in
urban transport systems. Their performance will be
described in depth, and some techniques to improve their
efficiency will be shown.

Key Issues in dc-electrified
system Consumption
In dc railway systems, the power consumed in traction
substations depends on the useful power required by the
trains. Useful power may be defined as the total amount
of power used by trains for traction and to make auxiliary
services work (air conditioning, lighting, etc.).
Figure 2 shows the Sankey diagram of a dc-electrified
railway system (López-López et al., 2014). The useful energy (consisting of traction energy plus auxiliary-service
energy) comes from the catenary (overhead contact line),
except a fraction of the auxiliary-service energy that
comes directly from the braking energy regenerated by the
train itself.

I E E E E l e c t r i f i c ati o n M agaz ine / september 2014

The energy absorbed for traction will transform into
kinetic energy once some portions are discounted, including train power chain losses, mechanical losses due to running resistance, train potential energy variations, and
auxiliary-equipment energy. If trains are equipped with
regenerative braking, a part of this energy may be given
back to the system, although losses due to other braking
systems, which are combined with the electric braking
(grouped as mechanical losses), have to be discounted.
Improving the design of the train may reduce these loss
sources, as shown in yellow in Figure 2. There are three
more loss sources, shown in red in Figure 2, that are system-infrastructure dependent.
xx
Substation losses: They represent losses in substation
transformers and rectifiers and depend on their power
rating. As will be shown later, if energy-storage systems
(ESSs) are used to improve the infrastructure, these losses may be reduced thanks to the load-shaving effect
they provide.
xx
Ohmic losses in conductors: They are proportional to conductor resistance and to the square of currents. Actions
such as increasing no-load voltage in substations may
reduce these losses. Other actions resulting in a shorter
average distance between generating and consuming
points in the line may also be beneficial.
xx
Rheostat losses: They are originated in momentary lack
of receptivity in the system, due to regenerated-energy
excess. This situation is more likely to occur in off-peak
hours (long headways) than in peak hours. Reversible

Table of Contents for the Digital Edition of IEEE Electrification Magazine - September 2014