- 250 -

ROAD SAFETY IN TUNNELS WITH

UNI- AND BI-DIRECTIONAL TRAFFIC IN AUSTRIA

Robatsch K., Nussbaumer C.

Institute for Traffic Engineering
Austrian Road Safety Board

ABSTRACT

The probability of an accident occurring and/or a motorist being killed in an accident is lower
in a tunnel than on other road sections. However, if an accident occur in a tunnel, the severity
of the injuries sustained is significantly higher than on other types of roads. This means, that
the cost per accident and the risk of being killed in an accident are higher in tunnels than on
motorways or expressways. In tunnel with unidirectional traffic, road safety is significantly
higher than in tunnels carrying bi-directional traffic. In tunnels with bi-directional traffic, the
accident rate is 28% and accident costs are 66% higher than in tunnels with unidirectional
traffic. The probability of being killed or involved in an accident is 65 to 71% higher in
tunnels with bi-directional traffic than in tunnels carrying unidirectional traffic.
Key words: tunnels, road safety, unidirectional, bi-directional

1. BACKGROUND
In recent years, a number of spectacular traffic accidents occurred in tunnels, which triggered
debates about the safety of road tunnels. Every year, an average of 71 accidents in motorway
and expressway tunnels occurs in Austria which causes an average of 15 fatalities, 38 severe
injuries and 91 minor injuries. The macroeconomic costs amount to a total of EUR 13 million.
The study “Tunnels with Uni- and Bi-directional traffic” of the Austrian Road Safety Board
by order of the Federal Ministry of Transport, Innovation and Technology (Robatsch K.,
Nussbaumer C., 2004) explores the traffic safety of road tunnels on motorways and
expressways compared with safety on other types of roads and also compares traffic safety in
tunnels carrying bi-directional traffic with safety in tunnels with unidirectional traffic.

The relative accident rates of all motorway and expressway tunnels studied are compared with
the corresponding rates for motorways and expressways as well as federal roads on open
sections. A detailed comparison of all motorway and expressway tunnels with unidirectional
and bi-directional traffic takes place. Moreover motorway and expressway tunnels of a
minimum length of one kilometre with unidirectional and bi-directional traffic are compared.

2. SAFETY IN TUNNELS VERSUS SAFETY ON OTHER TYPES OF ROADS

A variety of relative accident rates and the distance travelled in all of the tunnels studied are
compared with the corresponding figures for motorways, expressways and federal roads on
open sections.

In tunnels the accident rate and the casualty rate are significantly lower than on motorways,
expressways and federal roads on open sections. A comparison of the accident cost rates show
that cost rates of tunnels are higher than these of motorways, expressways and the open
sections of federal roads. Also, the fatality rate in tunnels is higher than on motorways and
expressways.

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 251 -

0,80 25,00
0,70
20,00
0,60
0,50 15,00
0,40
10,00
0,30
0,20 5,00
0,10
0,00 0,00
Accident cost rate Fatality rate
Accident rate Casualty rate

[Acc./1 mill. veh-km] [Cas./1 mill. veh-km] [EUR/1,000 veh-km] [Fat./1 bn veh-km]

Motorway Expressway Federal roads, open sections Tunnels

Figure 1: Relative accident rates for tunnels and other types of roads (2000-2001)

The probability of an accident occurring or road users being injured in a tunnel is lower than
on motorways, expressways and federal roads on open sections. But if an accident occurs the
injury severity in tunnels is especially high. Therefore the accident cost rate and the fatality
rate is higher in tunnels than on motorways and expressways.

Tunnels

Federal roads

Expressway

Motorway

0% 20% 40% 60% 80% 100%

fat sei nid sli

Figure 2: Seve rity of casualties in tunnels versus other types of roads (1999-2001)

The severity of casualties in tunnels is higher than on motorways, expressways and federal
roads on open sections. While 3.6% of casualties on motorways result in death, this rate in
tunnels is substantially higher at 10.6%. The proportion of those severely injured is slightly
higher in tunnels than on other types of roads.

3. SAFETY IN TUNNELS WITH UNI- AND BI-DIRECTIONAL TRAFFIC

This chapter compares accidents occurring in all studied tunnels with unidirectional and bi-
directional traffic on motorways and expressways.

3.1. Comparison of types of personal injury accidents

In tunnels with unidirectional traffic, same-direction accidents (including rear-end collisions
and accidents while changing lanes) account for about 57% of all accidents. In tunnels with
bi-directional traffic, the corresponding percentage is 45%. The main cause of rear-end
collision is the failure to maintain a safe distance to the vehicle in front. In tunnels with bi-
directional traffic, opposing direction accidents are the second most frequent type of accidents
at 41%. In tunnels with unidirectional traffic the second most frequent type of accidents are
single- vehicle accidents.

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 252 -

60,0

50,0

40,0

30,0

%
20,0

10,0

0,0
Single-vehicle Same direction Opposing direction Other accidents
accidents accidents accidents

Tunels with bi-directional traffic Tunnels with unidirectional traffic

Figure 3: Types of personal injury accidents of tunnels with bi-directional versus

unidirectional traffic in % (2001)

3.2. Selected assessment criteria

Table 8: Number, length and traffic intensity of tunnels with bi-directional and unidirectional
traffic (status 2000)
Tunnels with bi-directional Tunnels with unidirectional
traffic traffic
Number of tunnels studied 21 115
Total length [km] 87,709 101,357
Average length [km] 4,177 0,881
Traffic intensity [ADT] 12.704 8.958

In the calculations below, 21 tunnels with bi-directional traffic are compared with 115 tunnels
carrying unidirectional traffic. On average, a tunnel with bi-directional traffic is four to five
times as long as a tunnel with unidirectional traffic. At 12,704 vehicles per day, bi-directional
tunnels have more traffic intensity than unidirectional tunnels, which carry 8,958 vehicles per
day.

3.2.1 Comparison of relative accident rates

In the analysis below, a variety of relative accident rates have been calculated and compared
for tunnels with bi-directional traffic and unidirectional traffic. In addition to the absolute
accident figures and the relative accident rates it is helpful to also include the severity of
casualties. The calculations below comprise accident rates, accident cost rates, casualty rates,
fatality rates and involvement rates for accidents in tunnels with bi-directional and
unidirectional traffic.
0,30 30,00
0,25 25,00
0,20 20,00

0,15 15,00
0,10 10,00
0,05 5,00

0,00 0,00
Accident rate Casualty rate Involvement rate Accident cost rate Fatality rate

[Acc./1 mill. veh-km] [Cas./1 mill. veh-km] [Veh/1 mill. veh-km] [EUR/1,000 veh-km] [Fat./1 bn veh-km]

Tunels with bi-directional traffic Tunnels with unidirectional traffic

Figure 4: Relative accident rate for tunnels with bi-directional traffic and tunnels with
unidirectional traffic (1999-2001)

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 253 -

In tunnel with unidirectional traffic, the accident rate and the involvement rate are higher than
in tunnel with bi-directional traffic. The casualty rate is approximately equal in tunnels with
unidirectional and bi-directional traffic. The fatality rate and the accident cost rate are higher
in tunnels with bi-directional traffic than in tunnels with unidirectional traffic.

Tunnels with bi-directional traffic versus 1,60

tunnels with unidirectional traffic 1,39
1,40
1,28
1,20
1,03
1,00
0,78
0,80 0,70

0,60

0,40

0,20

0,00
Accident rate Accident cost rate Casualty rate Fatality rate Involvement rate
2001

Figure 5: Relative accident rates for tunnels with bi-directional traffic versus tunnels with
unidirectional traffic (tunnel with bi-directional traffic versus tunnels with unidirectional
traffic) (1999-2001)

It is not possible to draw a conclusion about the safety of tunnels with bi-directional traffic
and those with unidirectional traffic, as the relative accident rates have different
characteristics and other factors (such as traffic intensity, length of the tunnels) also have to
be taken into account. The impact of traffic intensity and tunnel length on relative accident
rates will be explored below.

3.2.2 Relationship between relative accident rates and traffic intensity

0,150

0,125
Accident rate

0,100

0,075
0,050

0,025

0,000
< 10.000 10.000-15.000 > 15.000

Vehicles/day

Figure 6: Average annual daily traffic (vehicles per day) and accident rate (personal injury
accidents per 1 million ve hicle-km) of tunnels with bi-directional and unidirectional traffic
(1999-2001)

Both in tunnels with bi-directional traffic and in tunnels with unidirectional traffic the
accident rate rises with increasing traffic intensity, as in tunnels with bi-directional traffic. In
tunnels with bi-directional traffic the risk of a head on collision and in tunnels with
unidirectional traffic the risk of a rear-end collision becomes higher.

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 254 -

3.2.3 Relationship between relative accident rates and tunnel length

1,500

1,250

1,000
Accident rate

0,750

0,500

0,250

0,000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Tunels with bi-directional traffic Tunnels with unidirectional traffic

Figure 7: Accident rate (personal injury accidents per 1 million vehicle-km) of tunnels with
bi-directional traffic and unidirectional traffic versus tunnel length (km) (1999-2001)

As shown in Figure 7 the length of the tunnel has a strong influence on the accident risk. Up
to a tunnel length of 1 kilometre the values are significantly higher than in longer tunnels.

Due to the major differences in length of tunnels with bi-directional traffic and unidirectional
traffic, the relevant data are not comparable. While 84 tunnels with unidirectional traffic are
less than one kilometre long, only two tunnels with bi-directional traffic fall into this
category. Therefore, further restrictions have to be introduced to enable a comparison of
tunnels with bi-directional traffic and unidirectional traffic.

Accident rate, casualty rate, involvement rate, fatality rate and accident cost rate depend
heavily on the length of the tunnel. This is particularly evident when comparing tunnels of
less than one kilometre length with longer tunnels. The highest risk for a tunnel accident to
occur is at the entrance area of a tunnel. At short tunnels this risk introduces more heavy in
the relative accident rates.

3.3. Selected assessment criteria for tunnels of more than one kilometre length
The length of a tunnel has a very substantial influence on relative accident rates. Particularly
tunnels of less than one kilometre length have very high accident rates. As the share of short
tunnels varies greatly, a comparison of safety is not possible. 73% of all tunnels with
unidirectional traffic (84) and 10% of all tunnels with bi-directional traffic (2) are shorter than
one kilometre. The question of whether tunnels with bi-directional traffic or tunnels with
unidirectional traffic are safer arisen, only with regard to longer tunnels, as short tunnels are
usually built as twin tube tunnels. For statistical reasons it seems meaningful to compare only
tunnels of a length of one kilometre and more.

The tables below compare 19 tunnels with bi-directional traffic and 31 tunnels with
unidirectional traffic after various relative accident rates. On average, tunnels with bi-

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 255 -

directional traffic that are longer than one kilometre are 2,3 times as long as tunnels with
unidirectional traffic.
0,25 30,00
0,20 25,00

0,15 20,00
15,00
0,10
10,00
0,05
5,00
0,00
0,00
Accident rat Casualty rate Involvement rate
Accident cost rate Fatality rate
[Acc./1mill. veh-km] [Cas./1mill. veh-km] [Veh/1mill. veh-km] [EUR/1,000 veh-km] [Fat./1bn veh-km]

Tunels with bi-directional traffic Tunnels with unidirectional traffic

Figure 8. Comparison of relative accident rates of tunnels of over 1 kilometre length with bi-
directional traffic and unidirectional traffic (1999-2001)

A comparison of tunnels of more than one kilometre length with bi-directional traffic and
unidirectional traffic shows that all of the relative accident rates selected are higher in tunnels
with bi-directional traffic than in tunnels with unidirectional traffic. The figures for tunnels
with bi-directional traffic include the Tauern tunnel accident (29th May 1999), which cause 12
fatalities, 4 severe injuries and 35 minor injuries. This accident is a rare catastrophic event,
but even when this accident is disregarded or given less weight, the relative accident rates for
tunnels with bi-directional traffic are still higher than those for tunnels with unidirectional
traffic.
1,71
1,80 1,69
1,66 1,65
Tunnels with bi-directional traffic versus

1,60
tunnels with unidirectional traffic

1,40
1,28
1,20

1,00

0,80

0,60

0,40

0,20

0,00
Accident rate Accident cost rate Casualty rate Fatality rate Involvement rate
2001

Figure 9: Relative accident rates in tunnels of over 1 kilometre length with bi-directional
traffic versus relative accident rates in tunnels with unidirectional traffic (tunnels with bi-
directional traffic versus tunnels with unidirectional traffic) (1999-2001)

4. SUMMARY
The probability of an accident occurring and a road user being injured is lower in a tunnel
than on motorways, expressways and federal roads on open sections. However, if an accident
occurs in a tunnel, the severity of the injuries sustained is significantly higher. This means,
that the accident cost rate and the fatality rate are higher in tunnels than on motorways or
expressways.

International Conference „Tunnel Safety and Ventilation“ 2004, Graz

 - 256 -

In comparison of tunnels of over 1 kilometre length, the accident cost rate and the
involvement rate are higher in tunnel with bi-directional traffic than in tunnels with
unidirectiona l traffic.

In a tunnel of more than one kilometre length with bi-directional traffic the probability of
being
• injured is 69% higher
• killed is 71% higher and
• involved in an accident is 65% higher
than in tunnels with unidirectional traffic.

Accidents in tunnels are not the main problem within the Austrian road network. The lacking
traffic morals regarding to maintain the speed limits and the failure to maintain a safe distance
between vehicles are fundamental problems. Every second tunnel accident can be attributed to
the short driving distance between vehicles. Instruments measuring driving distances between
vehicles as well as speed radar or section control equipment contribute to increased awareness
and a reduced accident risk. These measures have to be introduced in the entrance area of a
tunnel where the highest risk for accidents exists.

Since tunnel accidents are quite seldom and the reasons for their occurrence are not
exclusively due to the specific situation in a tunnel, problems with statistical significance of
accident data for tunnels are possible.

5. BIBLIOGRAPHY

Nadler, F. 2002: http://www.nast.at/nast/Daten_st.htm (Daten vom Sommer 2002).

Robatsch, K., Nussbaumer C.; 2004: Tunnels mit Gegenverkehr und Richtungsverkehr. In
Straßenforschung Heft 539, Bundesministerium für Wissenschaft und Verkehr (Hrsg.), Wien
2004.

International Conference „Tunnel Safety and Ventilation“ 2004, Graz