Page 284 - 360.revista de Alta Velocidad - Nº 5

Grande, Zacarías. Torralbo, Julia. Lobera, José Manuel. Sánchez-Cambronero, Santos. Castillo, Enrique.

1. Introduction and motivation

The design of high-speed railway lines could have a new alternative thanks to new methodologies
that offer important savings in construction and maintenance costs with no practical losses in
travel times.

This case study focuses on an alternate double-single track (ADST) methodology (Castillo, et
al., 2015). The main idea behind ADST consists of using single track where the infrastructure is
very expensive (tunnels and viaducts) and double track where it is cheaper (smooth orography),
if it is necessary.

The ADST methodology is especially suitable for peripheral sections where demand forecast is
low or intermediate. A double-track solution in these situations could lead to oversized lines
with inefficiency in exploitation and negative social impact investment financial returns. On
the other hand, the single-track would not satisfy passenger demand. The ADST performance in
peripheral lines is much closer to double- than to single-track performance, whereas the ADST
cost is much closer to single- than to double-track cost.
From the previous paragraph, the primacy of considering traffic volumes in rail design could be
deduced. An estimation of the demand can determine which the right solution for each case
study is: double-track line (high demand) or ADST line (low or intermediate demand).

The main tool required to develop an ADST line is an optimization program that allows us to
compare different track combinations and permits us to find the optimal sequence of single-
and double-track segments. Thus, construction costs are drastically reduced (up to 40%) and
maintenance costs are also substantially reduced (up to 50%) (Castillo, et al., 2015). Some
interesting publications related to the optimization of timetables are: (Amit & Goldfarb, 1971)
, (Burdett & Kozan, 2010), (Cacchiani & Toth, 2012), (Caprara, et al., 2002), (Carey & Crawford,
2007), (Castillo, et al., 2015), (Castillo, et al., 2011), (Castillo, et al., 2009), (Castillo, et al.,
2016), (D'Ariano, et al., 2007), (Pachl, 2014), (Sahin, 1999), etc.
Train routing and other optimization problems have been dealt with in (Assad, 1980), (Carey,
1994), (Carey & Lockwood, 1995), (Cordeau, et al., 1998), (D'Ariano & Pranzo, 2004), (Haghani,
1987), (Hellström, 1998), (Lin & Ku, 2013), (Ouyang, et al., 2009), (Petersen, et al., 1986),
(Yang & Hayashi, 2002), etc.

In the same way, the timetable must be optimized in order to reduce travel time. Since travel
times of different trains circulating along the network or line could be very different, and the
impact of a five-minute delay on a one-hour trip is not the same than on a three-hour trip, the
program uses relative travel times.
The relative travel time is the quotient:

Then, a relative time 1 means that we travel at maximum speed; contrary a relative time
value of 1.10 or 1.20 means that we have been used for the trip a 10% or a 20% more time,
respectively.
Delaying or advancing the departure or arrival time without changing the total travel times is
achieved forcing the trains to cross inside double-tracked segments.

282 360.revista de alta velocidad