Appendix 2 Risk Analysis Basic Example Summary

Document Sample
Appendix 2 Risk Analysis Basic Example Summary Powered By Docstoc
					Appendix 2: Risk Analysis Basic Example
Reference: UPTUN Report 5-4c, August 2006
Prepared by Inge Trijssenaar, TNO


Summary

Within UPTUN Workpackage 5 two examples of risk analysis are elaborated, a basic example
and a real road tunnel example. This appendix shows only the basic example. Each risk analysis
example consists of three modelling steps. Firstly, smoke dispersion calculations are performed
by INERIS and BRE using New Vendis and FASIT models, secondly evacuation calculations
are performed by BRE using CRISP model and finally the assessment is performed by TNO
using their model based on event tree analysis. This report describes the third part of the exam-
ples, which was performed by TNO.

The basic example tunnel is a two-lane bidirectional tunnel with a length of 1 km and a traffic
density of 30,000 vehicles per day. A risk analysis is performed for a baseline case and for sev-
eral upgrade cases with side exits, detection/alarm, sprinklers and all possible combinations of
these safety measures. Upgrading of the bi-directional tunnel requires temporary closure of the
tunnel, where the necessary closing time depends on the safety measure(s) to be installed. The
additional risk of a necessary detour during the closing time is included in the risk assessment of
the upgrade options.
The scenario’s included in the risk analysis are vehicle break-down/ damage–only accidents,
injury accidents, a small fire of 20 MW and a large fire of 100 MW. The TNO model uses the
CRISP results as input for the consequences of the fire scenario’s (number of injured persons
and number of casualties). Results are the expected value (number of victims per year), the in-
dividual risk (number of victims per person per km) and the societal risk (F/N curve). All results
are determined for as well injured as lethal victims. Figures 1 and 2 show the F/N curves for
lethal victims of the baseline and upgrade cases. The following effects of the safety measures
are observed: side exits decrease the number of victims, detection/alarm and suppression sys-
tems decrease the frequency of the higher numbers of victims. Similar trends are observed in the
F/N curves for injured persons. Figure 3 shows the expected value for the tunnel (the expected
number of lethal victims per year occurring in the tunnel). Furthermore figure 3 shows the rela-
tive contributions of injury accidents and fire accidents to this expected value and the effect of
the upgrade options to the (fire accident part of the) expected value. Figure 4 shows the effect of
the necessary detour during the closing time for upgrading. It is shown that detour risks are rela-
tively high compared to the risk of the tunnel alone, therefore upgrading should be combined
with necessary maintenance as much as possible. Limiting the time necessary for upgrading is
crucial for the net effect of the safety measures. Figure 5 shows the effect of the safety measures
on the individual risk. Based on the risk analysis of casualties and regarding the specifications
and assumptions of the basic example the following conclusions can be drawn for the basic ex-
ample tunnel1:
    • Taking detour into account, implementing detection & alarm is the best option with re-
         gard to expected value;
    • Side exits are most effective in reducing the maximum number of casualties in the so-
         cietal risk;

1
    N.B. the conclusions on the effectivity of safety measures are tunnel specific.



                                                                                                 1
                                        1,E+00
                 cumulative frequency


                                        1,E-02
                     [1/km/year]


                                                                                                          Baseline case
                                        1,E-04                                                            Case 1: side exits
                                                                                                          Case 2: detection/alarm
                                                                                                          Case 2a: side exits and detection/alarm
                                        1,E-06

                                        1,E-08

                                        1,E-10
                                                   1                   10                    100
                                                          number of lethal victims


                                        Figure 0-1: F/N curve for lethal victims- part I

                        1,E+00
cumulative frequency




                           1,E-02                                                                  Baseline case
    [1/km/year]




                                                                                                   Case 3: suppression
                           1,E-04                                                                  Case 3a: side exits, suppression


                           1,E-06                                                                  Case 3b: suppression and detection/alarm


                                                                                                   Case 3c: side exits, suppression and
                                                                                                   detection/alarm
                           1,E-08
                           1,E-10
                                               1                    10                     100
                                                       number of lethal victims

                                        Figure 0-2: F/N curve for lethal victims- part II
                                                                                                                                                                   Expected value [lethal victims per year]
                                                                                                                                                                                                                                                                                                                                                                                                             Expected value

                                                                                                                                          Ba                                                                                                                                                       Ca                                                Ba




                                                                                                                                                                               0,00E+00
                                                                                                                                                                               2,00E-02
                                                                                                                                                                               4,00E-02
                                                                                                                                                                               6,00E-02
                                                                                                                                                                               8,00E-02
                                                                                                                                                                               1,00E-01
                                                                                                                                                                               1,20E-01
                                                                                                                                                                               1,40E-01
                                                                                                                                                                               1,60E-01
                                                                                                                                                                               1,80E-01
                                                                                                                                                                               2,00E-01
                                                                                                                                                 se                                                                                                                                                     se                                                  se
                                                                                                                                                   lin                                                                                                                                                                                      Ca                 l   in
                                                                                                                                                      e                                                                                                                                                                                                               e




                                                                                                                                                                                                                                                                                                                                                                                       0,00E+00
                                                                                                                                                                                                                                                                                                                                                                                                  1,00E-02
                                                                                                                                                                                                                                                                                                                                                                                                               2,00E-02
                                                                                                                                                                                                                                                                                                                                                                                                                                        3,00E-02
                                                                                                                                                                                                                                                                                                                                                                                                                                                                   4,00E-02
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                       5,00E-02
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  6,00E-02

                                                                                                                                                                                                                                                                                                             2a                                  se
                                                                                                                                                                  ca                                                                                                                                                             Ca                                         ca
                                                                                                                                  C                                       se                                                                                                                                       :s
                                                                                                                                                                                                                                                                                                                     id             se                 1:                      s
                                                                                                                                      as
                                                                                                                                         e                                                                                                                                                                                e                 2:                sid                  e
                                                                                                                                                1:                                                                                                                                                                            ex                 de                    e
                                                                                                                    C                                 sid                                                                                                                                                                       it s                                        ex
                                                                                C                                       as                                   e                                                                                                                                                                                        te                         its
                                                                                    as                                                                                                                                                                                                                                                 an                   ct
                                                                                       e                                   e                                      ex
                                                                                                                                                                        its                                                                                                                                                                 d                  io
                                                                                           2a                                     2:                                                                                                                                                                         Ca                                 de                  n/
                                                                                                :s                                       de                                                                                                                                                                        se                                                    al
                                                                                                      id                                    t                                                                                                                                                     Ca                                                 te                       ar
                                                                                                           e                                    ec
                                                                                                                                                     ti o                                                                                                                                              se               3a             Ca               c   tio                    m
                                                                                                               ex                                        n/                                                                                                                                  Ca                               :s            se                    n/
                                                                                                                     i ts                                       al                                                                                                                                           3b                 up               3:
                                                                                                                            an                                    ar                                                                                                                              se               :s                                                  al
                                                                                                                                  d                                       m                                                                                                                            3c            up          pr             su arm
                                                                                                                                      de                                                                                                                                                                                            es             pp
                                                                                                                                         t   ec
                                                                                                                                                                                                                                                                                                            :s           pr
                                                                                                                                                                                                                                                                                                                           es          s io            re
                                                                                                                                                    ti o                                                                                                                                                         up                         n             ss
                                                                                                                                                        n/                                                                                                                                                          pr        s io
                                                                                                                                                                                                                                                                                                                                   n          &                 io
                                                                                                                            C                                al                                                                                                                                                        es                                          n
                                                                                           C                                    as                             ar
                                                                                                                                                                        m                                                                                                                                                si           &           sid
                                                                                               as                                  e                                                                                                                                                                                        on          de            e
                                                                                                  e                                      3:
                                                                                                                                                                                                                                                                                                                               &            te          ex
                                                                                                      3a                                        su                                                                                                                                                                                 si          ct              it s
                                                                                                               :s                                 pp                                                                                                                                                                                              io
                                                                            C                                       up                                      re                                                                                                                                                                        de             n/
                                                                                 as
                                                                                    e                                     pr                                   s   si                                                                                                                                                                     ex           al
                                                                                           3b                                  es                                       on                                                                                                                                                                   it s         ar
                                                                                                :s                                  si
                                                                                                                                         on                                                                                                                                                                                                       &             m
                                                                                                     up
                                                                                                           pr                                   &                                                                                                                                                                                                    de
                                                                            C                                   es                                   si                                                                                                                                                                                                  t. .
                                                                                as
                                                                                   e                                 si                                    de                                                                                                                                                                                                 .
                                                                                       3c                                 on                                     ex
                                                                                            :s                                  &                                     its
                                                                                                 up                                   de
                                                                                                    p                                      te
                                                                                                           re                                 c     t io
                                                                                                                ss                                         n/
                                                                                                                     io
                                                                                                                          n                                  al
                                                                                                                              &                                   ar
                                                                                                                                                                        m
                                                                                                                                  si
                                                                                                                                       de
                                                                                                                                              ex
                                                                                                                                                    its
                                                                                                                                                           &
                                                                                                                                                                d.
                                                                                                                                                                     ..




    Figure 0-4: expected value including detour [lethal victims per year]
                                                                                                                                                                                                                Figure 0-3: Expected value of tunnel, without detour [lethal victims/year]
                                                                                                                                                                                                                                                                                                                                                                                                                  accidents in tunnel
                                                                                                                                                                                                                                                                                                                                                                                                                  expected value of injury




3
                                                                                                                                                                                                  Tunnel only
                                                                                                                                                                                                                                                                                                                                                                                                                                                   Expected value tunnel incl. fires




                                                                                                                                                                                   50 km detour
                                                                                                                                                                                   25 km detour
                               2,50E-09
   lethalities /(person* km)



                               2,00E-09

                               1,50E-09

                               1,00E-09

                               5,00E-10

                               0,00E+00
                                              se



                                                            it s




                                                                                                             on
                                                                              m




                                                                                                                                it s
                                                                                               m
                                             ca



                                                           ex




                                                                                                                                                   ..
                                                                            ar




                                                                                                                                                                  ...
                                                                                                            si
                                                                                           ar




                                                                                                                              ex




                                                                                                                                                a.
                                                                        al




                                                                                                         es




                                                                                                                                                              &
                                          e



                                                       e




                                                                                         al




                                                                                                                                            n/
                                                                         n/




                                                                                                                           de
                                       lin



                                                      id




                                                                                                                                                             ts
                                                                                                       pr
                                                                                       n/




                                                                                                                                          io
                                                                     ti o
                                                      s




                                                                                                                                                            i
                                                                                                                        si
                                    se




                                                                                                     up
                                                                                     tio




                                                                                                                                                         ex
                                                                                                                                         ct
                                                   1:



                                                                   ec




                                                                                                                                      te
                                                                                                                       &
                                  Ba




                                                                                  ec



                                                                                                   s




                                                                                                                                                        de
                                                  e




                                                                                                                                    de
                                                                et




                                                                                                3:




                                                                                                                  n
                                              as




                                                                                 t




                                                                                                                                                    si
                                                                              de




                                                                                                                  io
                                                              d




                                                                                               e




                                                                                                                                   &
                                              C




                                                                                                              ss
                                                           2:




                                                                                                                                                  &
                                                                                           as
                                                                              d




                                                                                                                                n
                                                                                                             re
                                                                         an




                                                                                                                                              on
                                                           e




                                                                                                                              io
                                                                                           C
                                                      as




                                                                                                           pp



                                                                                                                           ss



                                                                                                                                            si
                                                                        ts
                                                      C




                                                                                                                                            s
                                                                                                       su



                                                                                                                         re
                                                                        i




                                                                                                                                         re
                                                                     ex




                                                                                                                       pp
                                                                                                       :




                                                                                                                                       pp
                                                                                                    3a
                                                                   de




                                                                                                                  su



                                                                                                                                    su
                                                                si




                                                                                                e



                                                                                                                   :
                                                                                               as



                                                                                                                3b




                                                                                                                                   :
                                                              :




                                                                                                                                3c
                                                           2a




                                                                                           C



                                                                                                            e



                                                                                                                              e
                                                          e




                                                                                                           as



                                                                                                                           as
                                                      as




                                                                                                       C



                                                                                                                        C
                                                      C




                                       Figure 0-5: individual risk (lethal victims per person per km)



Keywords
• Risk analysis example
• Probabilistic
• Long-term effectivity of safety measures
• Expected value
• F/N curve, societal risk
• Individual risk
Table of Contents

Appendix 2: Risk Analysis Basic Example                         1
1       Introduction                                            6
1.1     Risk analysis method                                     6
2       Basic example                                           8
2.1     System definition                                        8
2.1.1   Tunnel description: baseline case                        8
2.1.2   Upgrade options and detour                               8
2.2     Scenarios, damage and probabilities                      9
2.2.1   Scenario 0: vehicle break-down/ damage–only accidents    9
2.2.2   Scenario 1: injury accidents                             9
2.2.3   Scenario 2: small fire                                  10
2.2.4   Scenario 3: large fire                                  10
2.3     QRA results                                             13
2.3.1   F/N curves                                              13
2.3.2   Expected value                                          14
2.3.3   Expected value including detour                         15
2.3.4   Individual risk                                         16
3       Conclusions                                             18
3.1     Basic example                                           18
3.2     Overview effects of safety measures                     18
4       References                                              19
QRA results for injured victims                                 20




                                                                5
1        Introduction

Within UPTUN Workpackage 5 two examples of risk analysis are elaborated, a basic example
and a real road tunnel example. This appendix shows only the basic example. Each risk analysis
example consists of three modelling steps. Firstly, smoke dispersion calculations are performed
by INERIS and BRE using New Vendis model for the basic example and FASIT model for the
real road tunnel example. The FASIT model is selected for the real road tunnel example, be-
cause FASIT is more suitable than New Vendis for modelling the effect of a smoke extraction
system. Secondly evacuation calculations are performed by BRE using CRISP model and fi-
nally the assessment is performed by TNO using their model based on event tree analysis. The
steps for the basic example performed by INERIS and BRE are reported in [1]. This report de-
scribes the third step of the basic example, which was performed by TNO.


1.1      Risk analysis method
The applied method is a probabilistic quantitative risk analysis method. This method treats the
frequency and the consequence sides of the predicted risks with an equal emphasis. The risk
analysis considers a range of scenarios and includes the reliability of safety measures. Probabil-
istic quantitative risk analysis is a useful method to evaluate preventive safety measures on their
effect on the total risk.

Figure 1-1 shows the steps of the risk analysis method, which is applied for the UPTUN basic
example:
    • system definition
    • identification of accident scenarios
    • probability calculation
    • damage calculation (the damage calculations for the fire scenarios were performed by
        INERIS and BRE [1])
    • risk presentation
    • risk assessment
Figure 1-1 Risk Analysis [2]




                               7
2        Basic example

2.1      System definition


2.1.1    Tunnel description: baseline case
The basic example tunnel is a two-lane bidirectional tunnel with a length of 1 km. The tunnel is
used by 1250 vehicles per hour corresponding with a traffic density of 30,000 vehicles per day.
The traffic is assumed to consist of 81% cars and light goods vehicles (LGVs), 17% trucks and
2 % buses. An average occupation of 2.44 person per vehicle is assumed. A pressure difference
between the portals creates a natural longitudinal ventilation velocity from entry portal to exit
portal of 3 m/s. The baseline case is a “bare” tunnel without safety measures for improving fire
safety.


2.1.2 Upgrade options and detour

The baseline case is a “bare” tunnel without any safety measures for improving the fire safety.
Therefore several upgrade options are proposed to improve fire safety. The proposed safety
measures for upgrading include side exits, detection and alarm system, and suppression systems.
Suppression systems can be interpreted as techniques such as sprinklers, water mist technology
or isolation of the fire by means of inflatable tunnel plugs to starve the fire from oxygen and
separate smoke from people. All possible combinations of the safety measures are considered
and compared in the risk analysis. The following upgrade cases are defined:
Case 1: side exits
Case 2: detection and alarm
Case 2a: side exits and detection/ alarm
Case 3: suppression system
Case 3a: side exits and suppression system
Case 3b: suppression system and detection/alarm
Case 3c: side exits, suppression system and detection/alarm

The upgrade cases all result in decreasing damage, which is expressed as the number of victims.
The number of victims is a result of the CRISP evacuation model and input to the TNO risk
analysis model.

Side exits
The presence of side exits reduces the walking distance to safety for the persons in the tunnel.
The number of victims is input from the CRISP evacuation model. The users can evacuate by
tunnel portals and by safety exits integrated in the side of the tunnel every 200 m.

Detection and alarm
Users start evacuation at the same time due to alarm signal, the combination detection, alarm &
human response to the alarm (=start to evacuate) is assumed to have a failure frequency of 10%.

Suppression systems
It is assumed that suppression systems are able to extinguish a small fire of 20 MW. Further-
more they are able to reduce a large fire of 100 MW to a small fire of 20 MW. In case of failure,
the suppression systems do not succeed in extinguishing or reducing the fire. The failure fre-
quency of the suppression systems is assumed to be 5%.

Detour
Upgrading of the tunnel requires temporary closure of the tunnel, where the necessary closing
time depends on the safety measure(s) to be installed. The closing time of the bi-directional one
tube tunnel necessary for upgrading is:
Detection /alarm: 2 weeks during night time, a night is assumed to last 7 hours.
Suppression:       1 month complete closure
5 side exits:      12 months during the weekends (= 2 days per week)

Two detour options are calculated: a 25 km and a 50 km detour using secondary roads. The risk
of a necessary detour during the closing time is included in the risk analysis of the upgrade op-
tions. The frequency of injury accidents on secondary roads is 4,7*10-7 accidents per vehicle per
kilometre. This is almost five times higher as compared to a frequency of injury accidents in
tunnels, which is 1*10-7 accidents per vehicle per kilometre.

The safety measures are assumed to have a lifetime of 20 year, the risk of the detour is therefore
averaged over a 20 year period of tunnel-use. Note that for this basic example a constant traffic
density is assumed. Normally maintenance and upgrading are performed during off-peak hours.
Differences between peak and off-peak hours can be substantial.



2.2      Scenarios, damage and probabilities

The scenarios included in the risk analysis are:
• Scenario 0: vehicle break-down/ damage–only accidents;
• Scenario 1: injury accidents;
• Scenario 2: small fire (20 MW);
• Scenario 3: large fire (100 MW).


2.2.1 Scenario 0: vehicle break-down/ damage–only accidents

This scenario is limited to material damage to cars and sometimes to the tunnel wall. Fires are
not considered in the scenario itself. By definition, vehicle break-down and damage-only acci-
dents do not involve any victims, therefore it can not be found in the results of the risk assess-
ment. The scenario is included in the risk analysis as an initiating event for fire scenarios. The
accident frequency is 6 *10-6 accidents per vehicle per tunnel kilometre.


2.2.2 Scenario 1: injury accidents

“Injury accidents” are normal traffic accidents, where no fire occurs. Injuries can be caused by
the direct consequences of the collision (lethal or sublethal damage as a result of the impact).
The damage of an average injury accident is 1,16 injured victims and 0,036 lethal victims per



                                                                                                9
accident. The accident frequency of injury accidents is 1*10-7 accidents per vehicle per tunnel
kilometre.


2.2.3 Scenario 2: small fire

Scenario 2 is a small fire of 20 MW, a design fire scenario for which heat and smoke calcula-
tions are performed [1]. Input for the risk analysis is obtained from CRISP, the input comprises
the number of injured and lethal victims fallen during evacuation. CRISP calculates the Frac-
tional Effective Dose (FED) for each of the occupants at the end of the evacuation. Lethal vic-
tims have an FED of 100 percent (or more), injured victims are assumed to have an FED be-
tween 30 and 100 percent.
Injured persons from the injury accident itself may die during fire, because they are stuck in the
vehicle or not able to walk or to rescue themselves. It is assumed that 50% of the persons, who
are injured due to the collision, will die in a 20 MW fire. The remaining injured persons from
the injury accident are added to the CRISP results on injured persons.

For convenience, the event tree is divided in two parts for the small fire and large fire. Figure
2-1 shows the part of the event tree for the small fire scenario. The probabilities in the event tree
are specifically for the upgrade case 3c with side exits, suppression system and detection/alarm.
The small fire can be found in several branches of the event tree. For example, the small fire can
originate from a break-down of a car, followed by a fire that is neither extinguished nor de-
tected. Alternatively, it can originate from an injury accident followed by a fire. The break-
down of a car and the injury accidents are initiating events.

2.2.4 Scenario 3: large fire

Scenario 3 is a large fire of 100 MW.
Injured persons from the injury accident itself may die during fire, because they are stuck in the
vehicle or not able to walk (fast enough) or rescue themselves. It is assumed that all the persons,
who are injured due to the collision, will die in case of a large fire.
Figure 2-2 shows the event tree for the large fire scenario. The probabilities in the event tree are
again for the upgrade case 3c with side exits, suppression system and detection/alarm.
                                                                                         extinguish by
                     accident type                 fire             large fire           hand/ sprinkler    detection/ alarm
                                                                                            yes      0,95                            Car break-down or material damage accident
                                                     yes   0,0015    N.A.
                                                                                                                yes            0,9   Small fire detected
                                                                                            no       0,05
                                   0,797                                                                        no             0,1   Small Fire
                     car break-down
                     or material damage accident     no    0,9985                                                                    Car break-down or material damage accident

                                                                                           yes                                       Truck or bus break-down or material damage accident
                                                                      no         0,915
                                                                                                                yes            0,9   Small fire detected
                                                     yes   0,0015                           no       0,05
                                                                                                                no             0,1   Small fire

                                   0,187                              yes        0,085                                                 Scenario 3 large fire
                     truck or bus break-down
                     or material damage accident     no    0,9985                                                                    Truck or bus break-down or material damage accident
Initial accident
frequency                                                                                  yes       0,95                              Scenario 1 normal accident with injury
          6,10E-06
                                                                      no         0,915                          yes            0,9   Small fire detected
                                                                                            no       0,05
                                                                                                                no             0,1   Small fire
                                                     yes    0,01
                                                                      yes        0,085                                                 Scenario 3 large fire
                                 0,0164
                     accident with injury            no     0,99                                                                       Scenario 1 normal accident with injury




Figure 2-1: event tree of small fire scenario for upgrade case 3c (side exits, suppression system and detection/alarm)




                                                                                                                         11
                                                                             extinguish by hand/
                           Accident type             Haz. Mat.                    sprinkler            detection/alarm
                                                                                                       yes         0,9   Small fire, detected
                                                                                 yes        0,95
                                                                                                        no         0,1   Small fire

                                                    no haz.mat     1,000                               yes         0,9   Large fire, detected
Initial accident                                                                 no         0,05
frequency                      1,45E-10                                                                 no         0,1   Large fire
                   truck or bus break-down
                   or material damage accident       haz. mat      0,000                                                 Haz. Mat Scenario

                                                                                                       yes         0,9   Small fire, detected
                                                                                 yes        0,95
                                                                                                        no         0,1   Small fire

                                                        no         1,000                               yes         0,9   Large fire, detected
Initial accident                                                                 no         0,05
frequency                     8,50E-11                                                                  no         0,1   Large fire
                   accident with injury
                                                        yes                                                              Haz. Mat Scenario



Figure 2-2: event tree of large fire scenario for upgrade case 3c (side exits, suppression system and detection/alarm)
2.3           QRA results

Results of the probabilistic quantitative risk analysis are
• the expected value (number of victims per year)
• the individual risk (number of victims per person per km), and
• the societal risk (F/N curve).
All results are determined for as well injured as lethal victims. The results of the lethal victims
are described in this chapter, the results of the injured victims are given in chapter 5 of this ap-
pendix. The results of the injured victims show similar trends as those of the lethal victims.


2.3.1 F/N curves
Figure 2-3 and Figure 2-4 show the F/N curves for lethal victims of the baseline and upgrade
cases. The following effects of the safety measures are observed: side exits decrease the number
of victims, detection/alarm and suppression systems decreases the frequency of the higher num-
bers of victims. Similar trends are observed in the F/N curves for injured persons.


                              1,E+00
       cumulative frequency




                              1,E-02
           [1/km/year]




                                                                                       Baseline case
                              1,E-04                                                   Case 1: side exits
                                                                                       Case 2: detection/alarm
                                                                                       Case 2a: side exits and detection/alarm
                              1,E-06

                              1,E-08

                              1,E-10
                                        1                    10                  100
                                                number of lethal victims


                              Figure 2-3: F/N curve for lethal victims- part I




                                                                                                                 13
  cumulative frequency   1,E+00
                         1,E-02                                                  Baseline case
      [1/km/year]

                                                                                 Case 3: suppression
                         1,E-04                                                  Case 3a: side exits, suppression


                         1,E-06                                                  Case 3b: suppression and detection/alarm


                                                                                 Case 3c: side exits, suppression and
                                                                                 detection/alarm
                         1,E-08
                         1,E-10
                                  1                    10                  100
                                         number of lethal victims

                           Figure 2-4: F/N curve for lethal victims- part II



2.3.2 Expected value
Figure 2-5 shows the expected value for the tunnel, which is defined as the expected number of
lethal victims per year occurring in the tunnel. Furthermore Figure 2-5 shows the relative con-
tributions of injury accidents and fire accidents to this expected value and the effect of the up-
grade options to the (fire accident part of the) expected value. Figure 2-5 shows for the baseline
case that almost 80 percent of the expected value is caused by injury accidents and only slightly
more than 20 percent is caused by fire accidents.
                    6,00E-02

                    5,00E-02
   Expected value




                    4,00E-02                                                                                          Expected value tunnel incl. fires
                    3,00E-02
                                                                                                                      expected value of injury
                    2,00E-02                                                                                          accidents in tunnel

                    1,00E-02

                    0,00E+00
                                   e




                                                                                                                  n
                                                       m
                                           its
                                  s




                                                                                                              it s
                                                                                                               io




                                                                                                               m
                                                                                               su arm
                                                     ar
                               ca

                                         ex




                                                                                                         ss




                                                                                                         ar


                                                                                                             .
                                                                                                       ex




                                                                                                        t. .
                                                     al
                              e




                                                                                                      al
                                                                al

                                                                                                      re
                                       e

                                                n/




                                                                                                    de
                           in




                                                                                                     e
                                   sid




                                                                                                    n/
                                                            n/

                                                                                                  pp
                                                io




                                                                                                 sid
                          l
                       se




                                                          tio




                                                                                                 io

                                                                                                 &
                                             ct
                                  1:




                                                                                              ct
                     Ba




                                          te




                                                                                            it s
                                                                                             &
                                                        c




                                                                                           te
                                                     te

                                                                3:
                            se

                                        de




                                                                                           n




                                                                                         ex
                                                 de




                                                                                       de
                                                                                      s io
                          Ca




                                                          se
                                   2:




                                                                                     de
                                                                                     &
                                                d




                                                                                   es
                                                      Ca
                                           an
                                se




                                                                                  si
                                                                                  n
                                                                                pr

                                                                             s io
                             Ca




                                                                              &
                                         it s




                                                              up




                                                                           on
                                                                          es
                                       ex




                                                            :s




                                                                        si
                                                                        pr
                                                      3a
                                   e




                                                                      es
                                                                  up
                                 id




                                                                   pr
                                                     se
                               :s




                                                                :s

                                                                up
                                                Ca
                           2a




                                                            3b

                                                                     :s
                        se




                                                       se

                                                                  3c
                      Ca




                                                     Ca

                                                                se
                                                            Ca




                              Figure 2-5: Expected value of tunnel, without detour [lethal victims/year]



2.3.3 Expected value including detour
Figure 2-6 shows the effect of the necessary detour during the closing time for upgrading. It is
shown that detour risks are relatively high compared to the risk of the tunnel alone. This is due
to the -5 times- higher accident frequency on the roads used for the detour as well as the length
of the detour, which is 25 km compared to the 1 km tunnel length. So, while upgrading, the de-
tour causes a risk which is 125 times higher than the risk of use of the tunnel. After upgrading,
risk is reduced by the safety measures, however this risk reduction is not expected to pay off
against the risk increase during the upgrade. Especially the safety measure “side exits” does not
seem to pay off for the Basic Example, because of its relatively small effect on the expected
value combined with its long duration necessary for upgrading.
The conclusions on the effectivity of safety measures are specific for the characteristics as-
sumed for basic example tunnel, such as tunnel and detour lengths, one tube, ventilation, etc.
For instance for a tunnel with two tubes, a detour may not be necessary: while upgrading one of
the tube, traffic can use the other tube. Other assumptions influencing the outcome of the risk
assessment are the duration of the tunnel closure and the lifetime of the safety measures (20
year).
Because of the relatively high risks of the detour, upgrading should be combined with necessary
maintenance as much as possible. Limiting the time necessary for upgrading is crucial for the
effect of the safety measures.



                                                                                                                                                 15
  Expected value [lethal victims per year]
                                             2,00E-01
                                             1,80E-01
                                             1,60E-01
                                             1,40E-01
                                             1,20E-01                                                                                                                                    Tunnel only
                                             1,00E-01                                                                                                                                    25 km detour
                                             8,00E-02                                                                                                                                    50 km detour
                                             6,00E-02
                                             4,00E-02
                                             2,00E-02
                                             0,00E+00
                                                           se




                                                                                                                               on
                                                                           its




                                                                                                m




                                                                                                                                                 its
                                                                                                                m




                                                                                                                                                                    m
                                                                                           ar
                                                        ca




                                                                                                                                                                                    ..
                                                                          ex




                                                                                                            ar



                                                                                                                              si




                                                                                                                                                                   ar
                                                                                                                                              ex




                                                                                                                                                                                   d.
                                                                                         al




                                                                                                                             s
                                                                                                          al




                                                                                                                                                                 al
                                                       e




                                                                     e




                                                                                                                          re
                                                                                          n/




                                                                                                                                            de




                                                                                                                                                                                &
                                                                                                           n/
                                                    lin




                                                                                                                                                               n/
                                                                    sid




                                                                                                                       pp
                                                                                      ti o




                                                                                                                                                                             its
                                                                                                       ti o




                                                                                                                                                          t io
                                                                                                                                        si
                                                  se




                                                                                                                     su
                                                                                 ec
                                                                1:




                                                                                                                                                                           ex
                                                                                                    ec




                                                                                                                                        &




                                                                                                                                                           c
                                                 Ba




                                                                                                                                                        te
                                                                                  t
                                                                e




                                                                                                                                   on
                                                                                                                    3:
                                                                                                   t
                                                                               de




                                                                                                                                                                        de
                                                                                                de




                                                                                                                                                   de
                                                             as




                                                                                                                                   si
                                                                                                               e




                                                                                                                                                                      si
                                                                           2:
                                                           C




                                                                                                            as
                                                                                                d




                                                                                                                                                   &
                                                                                                                               es
                                                                                           an




                                                                                                                                                                    &
                                                                         e




                                                                                                                                              on
                                                                                                           C



                                                                                                                              pr
                                                                      as




                                                                                                                                                                 n
                                                                                         i ts




                                                                                                                          up




                                                                                                                                             si



                                                                                                                                                               io
                                                                    C




                                                                                                                                           es
                                                                                      ex




                                                                                                                                                              ss
                                                                                                                         :s



                                                                                                                                        pr



                                                                                                                                                         re
                                                                                                                     3a
                                                                                  e




                                                                                                                                      up




                                                                                                                                                          p
                                                                                 id




                                                                                                                                                       up
                                                                                                                   e
                                                                               :s




                                                                                                                                   :s
                                                                                                                as




                                                                                                                                                   :s
                                                                           2a




                                                                                                                                 3b
                                                                                                                C




                                                                                                                                                 3c
                                                                         e




                                                                                                                             e
                                                                      as




                                                                                                                                               e
                                                                                                                          as



                                                                                                                                            as
                                                                     C




                                                                                                                         C



                                                                                                                                        C


                                                        Figure 2-6: expected value including detour [lethal victims per year]



2.3.4 Individual risk

Figure 2-7 shows the effect of the safety measures on the individual risk. The individual risk
does not weigh the detour. Suppression systems show a relatively large effect on the individual
risk and on the expected value. Detection/ alarm has a relatively small effect on individual risk
and expected value of the tunnel.
                                                                                                                                                                                                lethalities /(person* km)


                                                                                                                                                   Ba
                                                                                                                                                        se




                                                                                                                                                                                         0,00E+00
                                                                                                                                                                                                    5,00E-10
                                                                                                                                                                                                               1,00E-09
                                                                                                                                                                                                                          1,50E-09
                                                                                                                                                                                                                                     2,00E-09
                                                                                                                                                                                                                                                2,50E-09



                                                                                                                                                                 lin
                                                                                                                                                                       e
                                                                                                                                                                           ca
                                                                                                                                          C                                       se
                                                                                                                                              as
                                                                                                                                                   e
                                                                                                                                                        1:
                                                                                    C                                  C                                   s     id
                                                                                        as
                                                                                             e
                                                                                                                            as                                         e
                                                                                                 2a                              e                                         ex
                                                                                                                                      2:                                          it s
                                                                                                    :   si                               d
                                                                                                              de                               et
                                                                                                                                                       ec
                                                                                                                   ex                                       ti o
                                                                                                                      i   ts                                    n/
                                                                                                                                 an                                    al
                                                                                                                                                                            ar
                                                                                                                                      d                                       m
                                                                                                                                          de
                                                                                                                                             t     ec
                                                                                                                                                        tio
                                                                                                                                                                n/
                                                                                                                                 C                                al
                                                                                                 C                                   as                                   ar
                                                                                                     as                                   e                                    m
                                                                                                          e                                    3:
                                                                                                              3a                                  s
                                                                                                                 :   su
                                                                                                                                                        up
                                                                                    C                                       pp
                                                                                                                                                                pr
                                                                                                                                                                  es
                                                                                        as
                                                                                             e                                re                                           si
                                                                                                                                                                                on
                                                                                                  3b                                  ss
                                                                                                     :    su
                                                                                                                                           io
                                                                                                                                                   n
                                                                                                               pp                                      &
                                                                                    C                            re                                         si
                                                                                                                                                                 de
                                                                                        as                                ss
                                                                                             e                                   io                                    ex
                                                                                                 3c                                   n                                        it s
                                                                                                    :     su                              &
                                                                                                                                               de
                                                                                                               pp                                te
                                                                                                                     re                                    ct
                                                                                                                        s   si                                  io
                                                                                                                                                                     n/
                                                                                                                                 on                                       a.
                                                                                                                                      &                                      ..
                                                                                                                                              si
                                                                                                                                                   de
                                                                                                                                                        ex
     Figure 2-7: individual risk (lethal victims per person per kilometre tunnel)




                                                                                                                                                           i     ts
                                                                                                                                                                       &
                                                                                                                                                                           ...




17
3        Conclusions

3.1      Basic example
A probabilistic quantitative risk analysis is performed for the UPTUN basic example.
Upgrading of the bi-directional tunnel requires temporary closure of the tunnel, where the nec-
essary closing time depends on the safety measure(s) to be installed. The additional risk of a
necessary detour during the closing time is included in the risk assessment of the upgrade op-
tions. The detour risks are relatively high compared to the risk of the tunnel alone, therefore up-
grading should be combined with necessary maintenance as much as possible. Limiting the time
necessary for upgrading is crucial for the effect of the safety measures.
Based on the risk analysis of casualties and regarding the specifications and assumptions of the
basic example the following conclusions can be drawn for the basic example tunnel:
    • Taking detour into account, implementing detection & alarm is the best option with re-
         gard to expected value;
    • Side exits are most effective in reducing the maximum number of casualties in the so-
         cietal risk;
The conclusions on the effectivity of safety measures are specific for the characteristics as-
sumed for basic example tunnel, such as dimensions, one tube, ventilation, etc. The conclusions
should not be extrapolated to other tunnels, instead a risk analysis should be performed to assess
the effectivity of safety measures in these tunnels.

3.2      Overview effects of safety measures

Table 3.1 gives an overview of the effect of safety measures based on the risk analysis of the
basic example tunnel and the real road tunnel example. Please note that the results are specific
for these tunnels.


Table 3.1Overview effects of safety measures
                      Expected value              Individual risk          Societal risk
                      tunnel       Including      tunnel                   Frequency Number of
                                   detour                                                victims
Basic example (tunnel specific results!)
Side exits                +            --                    +                  ±             +
Detection/ alarm          +             ±                    +                  +             ±
Suppression              ++             -                   ++                  +             ±
4        References

[1] “Risk assessment integrating the coupling between the smoke dispersion models and
    evacuation model”, UPTUN Workpackage 5 report.
[2] SafeT Workpackage 1 report, “Current state of practice”, I.J.M. Trijssenaar et al., April
    2005.
[3] Communication with Jeremy Fraser-Mitchell from BRE, UK and Dominic Walley, CEBR,
    UK.




                                                                                          19
5                          QRA results for injured victims




                                        1,E+00
                 cumulative frequency




                                         1,E-02                                                           Baseline case
                     [1/km/year]




                                                                                                          Case 1: side exits

                                         1,E-04                                                           Case 2: detection/alarm
                                                                                                          Case 2a: side exits and detection/alarm


                                         1,E-06

                                         1,E-08

                                         1,E-10
                                                    1                   10                   100
                                                        Number of injured victims


                                        Figure 5-1: F/N curve for injured victims- part I



                               1,E+00
                                                                                                   Baseline case
    cum u lative frequ en cy




                               1,E-02                                                              Case 3: suppression
          [1/km /year]




                                                                                                   Case 3a: side exits, suppression
                               1,E-04
                                                                                                   Case 3b: suppression and detection/alarm
                               1,E-06
                                                                                                   Case 3c: side exits, suppression and
                                                                                                   detection/alarm
                               1,E-08


                               1,E-10
                                            1                    10                   100
                                                    Number of injured victims


                                        Figure 5-2: F/N curve for injured victims- part II
                                                                                                                                                                                                                                   Expected value


                                                                                                                                                                Ba
                                                                                                                                                                     se




                                                                                                                                                                                                  0,00E+00
                                                                                                                                                                                                             2,00E-01
                                                                                                                                                                                                                        4,00E-01
                                                                                                                                                                                                                                    6,00E-01
                                                                                                                                                                                                                                               8,00E-01
                                                                                                                                                                                                                                                          1,00E+00
                                                                                                                                                                                                                                                                     1,20E+00
                                                                                                                                                                                                                                                                                1,40E+00
                                                                                                                                                                                                                                                                                           1,60E+00
                                                                                                                                                                                                                                                                                                      1,80E+00



                                                                                                                                                                       lin
                                                                                                                                                                          e
                                                                                                                                                                                    ca
                                                                                                                                                   C                                       se
                                                                                                                                                        as
                                                                                                                                                           e
                                                                                                                                                                 1:
                                                                                               C                                      C                                  sid
                                                                                                   as                                      as                                   e
                                                                                                      e                                       e                                     ex
                                                                                                          2a                                       2:
                                                                                                                :s                                                                         it s
                                                                                                                      id                                de
                                                                                                                           e                               te
                                                                                                                               ex                             c      tio
                                                                                                                                  i   ts                                   n/
                                                                                                                                            an                                al
                                                                                                                                               d                                 a       rm
                                                                                                                                                    de
                                                                                                                                                            te
                                                                                                                                                               c    tio
                                                                                                                                                                       n/
                                                                                                                                            C                            al
                                                                                                          C                                     as                               ar
                                                                                                               as                                  e                                m
                                                                                                                  e                                     3:
                                                                                                                      3a                                        su
                                                                                              C                                :s                                 pp
                                                                                                  as                                up                                     re
                                                                                                     e                                     pr                                ss
                                                                                                         3b                                   e   ss                                   io
                                                                                   C                           :s                                      io                                    n
                                                                                       as                           up                                      n
                                                                                          e                              pr                                     &
                                                                                              3c                            e   ss                                  si
                                                                                                   :s                                 io                                 de
                                                                                                        up
                                                                                                           p                               n                                    ex
                                                                                                               re                              &                                       its
                                                                                                                 ss                                de
                                                                                                                      io                             te
                                                                                                                           n                           ct
                                                                                                                                                                    io
                                                                                                                               &                                       n
                                                                                                                                    si                                     /a
                                                                                                                                         de                                     la
                                                                                                                                                ex                                 r   m
                                                                                                                                                   i   ts
                                                                                                                                                            &
                                                                                                                                                                 de
                                                                                                                                                                   te
                                                                                                                                                                     ct
                                                                                                                                                                                 i. .
     Figure 5-3: Expected value of tunnel, without detour [injured victims/year]                                                                                                     .




21