Fixed Fire Fighting in Road Tunnels Current Practices and Recommandations

www.piarc.
org
2016R03EN
FIXED FIRE FIGHTING

SYSTEMS IN ROAD TUNNELS:
CURRENT PRACTICES
AND RECOMMENDATIONS
Technical Committee C.3.3 Road Tunnel Operations
of the World Road Association
STATEMENTS
The World Road Association (PIARC) is a nonprofit organisation established in 1909 to improve international
co-operation and to foster progress in the field of roads and road transport.
The study that is the subject of this report was defined in the PIARC Strategic Plan 2012 2015 and approved
by the Council of the World Road Association, whose members are representatives of the member national
governments. The members of the Technical Committee responsible for this report were nominated by the
member national governments for their special competences.
Any opinions, findings, conclusions and recommendations expressed in this publication are those of the authors
and do not necessarily reflect the views of their parent organisations or agencies.
This report is available from the internet site of the World Road Association (PIARC)
http://www.piarc.org
Copyright by the World Road Association. All rights reserved.
World Road Association (PIARC)

Tour Pascal B - 19e tage
92055 La Dfense cedex, FRANCE
International Standard Book Number 978-2-84060-375-7

Frontcover Marina Coastal Expressway, Singapore
FIXED FIRE FIGHTING
SYSTEMS IN ROAD
TUNNELS: CURRENT
PRACTICES AND
RECOMMENDATIONS
Technical Committee 3.3 Road Tunnel Operations
of the World Road Association
2016R06EN FIXED FIRE FIGHTING SYSTEMS IN ROAD TUNNELS
AUTHORS/ACKNOWLEDGMENTS
This report has been prepared by Working Group 4 of the Technical Committee 3.3 of the World
Road Association (PIARC).
The contributors to the preparation of this report are:
Radim Bajger (Czech Republic) Norris Harvey (USA)

Matthew Bilson (Australia) Haukur Ingason (Sweden)
Grzegorz Blaszczyk (Poland) Marko Jarvinen (Finland)
Rune Brandt (Sweden / Switzerland) Roland Leucker (Germany)
Lionel Brown (UK) Ulf Lundstrm (Sweden)
Harald Buvik (Norway) Toshiro Otsu (Japan)
Ricky Carvel (UK) Xavier Ponticq (France)
Gary Clark (UK) Norman Rhodes (UK)
Bruce Dandie (Australia) Marien Riemens (Netherlands)
Ignacio Del Rey (Spain) Juan Manuel Sanz (Spain)
Arnold Dix (Australia) Peter Sturm (Austria)
Leslie Fielding (UK) Fathi Tarada (UK)
Sylvain Garnier (France) Pauli Velhonoja (Finland)
Robin Hall (UK)
Reviews were provided by Wah Onn Adrian Cheong (Singapore), Boon Hui Chiam (Singapore),
Gary English (USA), Jorgen Holst (Denmark), Ryu Ji Hyun (South Korea), Nam-Goo Kim
(South Korea), Ronald Mante (Netherlands), Frederic Walet (France), Urs Welte (Switzerland),
Franz Zumsteg (Switzerland). Additional reviews provided by Bernhard Kohl on behalf of BASt
and Dirk Sprakel on behalf of COSUF.
The Working Group was chaired by Leslie Fielding (UK) and Bruce Dandie (Australia) was the
Secretary.
The English editors of this report were Bruce Dandie (Australia) and Norris Harvey (USA). The
translation into French of the original version was undertaken by Sylvain Garnier (France) and
Xavier Ponticq (France).
This report was developed over two PIARC cycles. During the 2008-2011 cycle, the working
group was led by Arnold Dix (Australia).
The Technical Committee was chaired by Ignacio Del Rey (Spain) with Marc Tesson (France),
Fathi Tarada (United Kingdom), Juan Marcet (Argentina) respectively being the French, English
and Spanish speaking secretaries.
EXECUTIVE SUMMARY
2016R03EN
FIXED FIRE FIGHTING SYSTEMS

IN ROAD TUNNELS: CURRENT PRACTICES
AND RECOMMENDATIONS
Fixed Fire Fighting Systems (FFFS) have been routinely used in road tunnels in countries such
as Japan and Australia for decades, and there is increased interest in the use of FFFS in parts of
Europe, North America and Asia.
Large fire events in road tunnels continue to show the consequences of such an event. These
consequences can be to the users, the tunnel infrastructure, and the impact to the wider road
network on society. This has maintained the pressure for further improvements to techniques
and technologies to manage the risk and consequence of fires in road tunnels. FFFS are
increasingly seen as a method that can deliver user safety and infrastructure protection, and can
be used as a risk reduction measure. However, their use is not widespread for various political,
economic, technical and social reasons. It is still recognised that FFFS may not be the most
appropriate measure to adopt in all circumstances or in all locations.
Within this report, the functional impact FFFS can make to the performance of tunnel fire safety
systems is discussed. Information is presented about the types of systems available, their use in
road tunnels for various countries, and advice provided on the design and selection of appropriate
FFFS. Where FFFS are adopted, it is essential that they are correctly designed, installed,
integrated, commissioned, maintained, tested and operated. Where FFFS are installed, it is
recommended that they are activated in the early stages of a fire to minimise fire growth and to
provide the desired effectiveness.
I
II
FIXED FIRE FIGHTING SYSTEMS IN ROAD TUNNELS 2016R03EN
CONTENTS
1. INTRODUCTION............................................................................................................................................................3
1.1. PURPOSE OF THIS REPORT..........................................................................................................................3

1.2. SCOPE OF THE REPORT................................................................................................................................3
1.3. TARGET GROUP...............................................................................................................................................4
1.4. THE ROLE OF FIXED FIRE FIGHTING SYSTEMS...................................................................................4
2. PREVIOUS WORK.........................................................................................................................................................6
2.1. PIARC PUBLICATIONS..................................................................................................................................6

2.2. OTHER REPORTS AND GUIDANCE............................................................................................................8
3. DECISION FACTORS....................................................................................................................................................9
3.1. INTRODUCTION..............................................................................................................................................9
3.2. COMPLIANCE WITH REGULATIONS, GUIDELINES AND LEGAL CONSIDERATIONS................9
3.3. RISK ASSESSMENT.........................................................................................................................................9
3.4. LIFE SAFETY..................................................................................................................................................10
3.5. ASSET PROTECTION....................................................................................................................................12
3.6. ADDITIONAL TRAFFIC REGIMES............................................................................................................13
3.7. FIRE-FIGHTING RESPONSE.......................................................................................................................13
3.8. OPERATIONS, MAINTENANCE AND REGULAR TESTING ...............................................................14
3.9. COST BENEFIT CONSIDERATIONS..........................................................................................................14
3.10. SUSTAINABILITY .......................................................................................................................................16
3.11. LEGAL CONSIDERATIONS.......................................................................................................................16
4. DESIGN CONSIDERATIONS ................................................................................................................................... 17
4.1. RISK ASSESSMENT.......................................................................................................................................18

4.2. TYPE OF SYSTEM.........................................................................................................................................19
4.3. WATER DISCHARGE CHARACTERISTICS (MIST OR DELUGE) ......................................................19
4.4. WATER SUPPLY .............................................................................................................................................20
4.5. DRAINAGE......................................................................................................................................................21
4.6. SPACE CONSIDERATIONS..........................................................................................................................21
4.7. FIRE DETECTION AND ACTIVATION STRATEGY................................................................................21
4.8. TUNNEL ENVIRONMENT...........................................................................................................................22
4.9. SYSTEM INTEGRATION..............................................................................................................................23
4.10. INTERACTION OF FFFS WITH VENTILATION....................................................................................24
4.11. FURTHER ASPECTS....................................................................................................................................28
5. SYSTEM DEFINITION / PROCUREMENT..........................................................................................................28
5.1. SYSTEM DEFINITION..................................................................................................................................28

5.2. TESTING AND COMMISSIONING.............................................................................................................33
5.3. APPROVALS....................................................................................................................................................33
5.4. TRAINING........................................................................................................................................................33
5.5. COST.................................................................................................................................................................33
6. RESEARCH AND ANALYSIS...................................................................................................................................34
6.1. RESEARCH PROGRAMS..............................................................................................................................34

6.2. MODELLING ..................................................................................................................................................35
7. CONCLUSIONS AND RECOMMENDATIONS....................................................................................................36
7.1. CONCLUSIONS..............................................................................................................................................36
7.2. RECOMMENDATIONS..................................................................................................................................36
7.3. FUTURE WORK..............................................................................................................................................37
8. BIBLIOGRAPHY / REFERENCES..........................................................................................................................38
9. GLOSSARY....................................................................................................................................................................43
10. APPENDICES..............................................................................................................................................................44
APPENDIX 1 - QUESTIONNAIRE DATA..........................................................................................................44

APPENDIX 2 - TYPES OF SYSTEMS................................................................................................................58
APPENDIX 3 - MAINTENANCE AND TESTING............................................................................................68
APPENDIX 4 - RESEARCH AND EVALUATION PROGRAMS....................................................................71
APPENDIX 5 - MODELLING...............................................................................................................................75
APPENDIX 6 - SUSTAINABILITY.....................................................................................................................77
3
1. INTRODUCTION
1.1. PURPOSE OF THIS REPORT
This report replaces the previous 2008 PIARC guidance on Fixed Fire Fighting Systems (FFFS)
Road Tunnels: An Assessment of Fixed Fire Fighting Systems [30]. It advances the understanding
of FFFS, their application, benefits and limitations, as well as providing advice on the design,
procurement installation, testing, commissioning, maintenance and operation of FFFS. The
report also looks at international experience based on current installations, test programmes and
real life incidents.
Since the publication of the 2008 report, a greater understanding of FFFS has expanded their use
and application around the world. This has occurred as a consequence of test and research
programmes, incident data reports from existing FFFS, and the development of designs for new
and existing tunnels.
In some countries, risk and cost benefit analyses are used to consider the application of FFFS as
a measure to assist in making infrastructure both safer, and more durable in the event of an
incident. However, for various political, economic, technical, and social reasons, it is recognised
that FFFS may not be the most appropriate measure to adopt in all circumstances. These reasons
can include where a road tunnel has a dedicated fire service to provide a similar response in a
timely manner, where government directive asserts that FFFS will not be applied in that particular
countrys tunnels, or where FFFS will not be maintained and operated to the degree of reliability
and availability required. Where FFFS are installed, it is essential that they be correctly designed,
installed, integrated, commissioned, maintained, tested, and operated with a high level of
reliability and availability, so that the systems are available for use as required.
The purpose of this report is to provide decision makers and designers with information to assist
them with their understanding of the parameters of FFFS, and to provide guidance on whether
or not to include FFFS in their road tunnels.
1.2. SCOPE OF THE REPORT
The scope of this report is to provide strategic guidance and advice on FFFS to allow their
implementation to be considered in a balanced manner. In doing this, the report provides
information on the application of FFFS, including the design, procurement, installation,
integration, testing, commissioning, operation and maintenance of water-based FFFS in tunnels.
The report consolidates current and previous work by PIARC and other organisations, and
includes data from numerous full-scale fire tests where data were available.
The term FFFS refers to a range of technologies that use water as the suppression agent, or water
with an additive or some other extinguishing agent. These systems are installed as part of the
tunnel infrastructure and require no additional elements to be added when called upon to fight
fires. As such, these systems are part of the fixed installation, having been installed for the
specific purpose of controlling a fire incident over a specific area and are activated automatically,
semi-automatically, or manually from a remote location. Therefore, the term FFFS does not
apply to manual fire-fighting methods that only have a portion of the infrastructure fixed, such
as the case with standpipes and hydrants.
4
1.3. TARGET GROUP
The Target Group for this report is road authorities, tunnel owners, tunnel operators, and tunnel
managers. This report is also intended to be of benefit to designers, emergency services,
contractors, tunnel safety officers, and the industry in general.
1.4. THE ROLE OF FIXED FIRE FIGHTING SYSTEMS
The role of FFFS for road tunnels is to provide facilities for tunnel owners and operators to assist
with the early suppression and subsequent management of fires. In this manner, the consequences
of a fire event to tunnel users, the tunnel infrastructure, and the societal impact due to disruption
to the wider road network can be mitigated. Their installation provides the fixed infrastructure
within a tunnel to enable fires to be addressed more quickly and more easily than if incident
responders had to provide and deliver alternate systems to the fire site to respond to the event by
other means.
As FFFS are part of the tunnel infrastructure, they allow fire control to be initiated from a
remote location automatically, semi-automatically, or manually. This provides advantages in that
FFFS allow:
fires to be addressed in a timely manner, even before the fire brigade arrives at the incident site;
delivery of sufficient water to the fire site, such that control or suppression of a fire can occur
before the fire develops into a full scale conflagration;
the fire brigade to manage the fire incident without putting themselves at risk by being in the
near vicinity of a fire; and
the fire brigade to fully extinguish the fire once it has been suppressed (if it has not already
been extinguished).
As a fire is able to be addressed in a timely manner, the operation of FFFS should occur while
the fire is still in the growth phase. This permits FFFS to be more effective than if they were
applied to a fire when it had developed to a full scale conflagration. By applying water directly
to the fire site, FFFS retard the growth rate of the fire by cooling the surface of the burning
material and inhibiting the combustion process. The end result is that the consequence of a fire
is reduced, enhancing life safety and reducing damage to the tunnel structure and tunnel
equipment. FFFS also provide more tenable conditions for the fire brigade to access and fully
extinguish the fire.
Properly designed, installed, integrated, commissioned, maintained, tested and operated FFFS
will:
provide early suppression and control of a fire event;

retard the fire growth rate, thereby inhibiting the combustion process and reducing the heat
output;
remove heat from the environs of the fire by cooling the surrounding area during an incident;
limit the potential for fire to spread between vehicles;
extend the available escape time for tunnel users;
5
improve overall tenability for fire fighters, enabling them to respond to the event more
effectively;
reduce the likelihood and extent of structural damage;
limit the severity and extent of damage to tunnel systems and equipment;
allow the asset to return to service in a shorter period of time following a fire; and
return the external road network to full integrity in a shorter period of time following a fire.
Water based FFFS in tunnels can be divided into Deluge Systems and Water Mist Systems, both
with and without the use of foam additives. However, water based systems without additives
represent the vast majority of installed systems. Even though Deluge Systems and Water Mist
Systems use valves to enable operation over a discrete zone of the tunnel, the terms Deluge
Systems and Water Mist Systems having the predominant features as described in chapters 1.4.1
and 1.4.2, will apply herein.
Where FFFS are installed, it is essential that they be correctly designed, installed, and integrated
into the tunnel system, as well as properly tested, commissioned, maintained, and operated.
Where installed, it is recommended that activation should occur in the early stages of a fire to
minimise fire growth and ensure the desired effectiveness.
1.4.1. Deluge Systems
Deluge Systems are typified by a zoned water application, characterised by a significant

proportion by volume of relatively large water droplets. With Deluge Systems, the size of the
water droplet counters the extent of drift that may occur with longitudinal ventilation systems.
The exact performance of these systems varies from tunnel to tunnel as their performance is
usually specified as an application rate over a discrete section of tunnel, or as a delivered density
application rate in mm/min or l/min/m2, and not on the basis of droplet size distribution
(appendixA2.2, page58 for a more detailed description). Deluge Systems predominantly control
a fire by removing heat and inhibiting the combustion process by cooling surfaces directly at and
adjacent to the fire site.
1.4.2. Water Mist Systems
Water Mist Systems can be either low or high pressure, however, the pressures used are typically
higher than those used for Deluge Systems. Water Mist Systems are typically used where the
volume of water, spatial considerations, or weight restrictions, are an issue. Water Mist Systems
are characterised by relatively fine water droplets, which assist cooling by the evaporative
process. Some droplets get entrained to the seat of the fire by air convection, while others reach
the fire from above due to the high momentum imparted to them. Also, due to the high momentum,
drift along the tunnel due to longitudinal ventilation systems or external ambient pressure effects
is generally minor. Mist systems are specified based on the volume of the tunnel in the application
zone in l/min/m3 (appendix A2.3, page63 for a more detailed description). Water Mist Systems
predominantly operate by gas cooling and provide limited cooling of surfaces immediately at
and adjacent to the fire site.
6
1.4.3. Common Features and Variances
Both systems are characterised by:
a water supply with sufficient reliability, quality, quantity and pressure for application at the
required rate over the designed tunnel area;
an activation valve (typically a deluge valve, section valve, or a solenoid) that controls the flow
of water to the distribution network and hence does not rely on localised fusible link sprinkler
heads;
a water distribution network between the activation valve and the spray nozzle;
the ability to deliver a predefined volume of water over the designated fire zone for a
predetermined period of time.
Water Mist Systems vary from Deluge Systems in that Water Mist Systems typically:
use higher pressures than Deluge Systems;

use smaller diameter pipework than Deluge Systems;
uses less water volumes and flow rates for the same area of coverage;
use more specialised material and equipment such as for pumps, pipes and nozzles due to
the higher operating pressure, and the requirement to keep the fine spray nozzles clear of
any particles that may occur in the pipe network and block the nozzle openings. The need to
eliminate blockages may also require the addition of filtration systems.
2. PREVIOUS WORK
The use and application of FFFS in tunnels has been debated for many years both within PIARC
and other eminent organisations. In recent years, the situation has changed from a position where
FFFS were not recommended, to that where FFFS are noted as having benefits both during a fire
event and in returning the asset to service following a fire event. This change in position is
reflected in the increasing volume of documentation and guidance that has arisen as technical
knowledge of FFFS has improved by actual applications; a significant number of large scale fire
tests and test programmes; and incident data reports on the effectiveness of properly designed,
installed, integrated, commissioned, maintained, tested and operated FFFS.
2.1. PIARC PUBLICATIONS
Publications of the World Road Association (PIARC) on FFFS over the years are summarised
below.
2.1.1. 1999 - Fire and Smoke Control in Road Tunnels
In 1999 the PIARC publication Fire and Smoke Control in Road Tunnels [29] chapter VI.3.3
Water Supply, and chapter VI.3.4 Sprinklers, reviewed previous work by PIARC and made
recommendations for the provision of a fixed water supply and the use of sprinklers. It
recommended that water supply standpipes be provided for tunnels greater than 200 m and that
hydrants for fire brigade use be placed at a spacing of 100 to 200 m (refer chapter VI.3.3.3).
However, it also noted that sprinklers were not commonly used because most fires start in
the motor room or in the compartment, and sprinklers are of no use till the fire is open. Sprinklers
7
can be used, however, to cool down vehicles, to stop the fire from spreading to other vehicles
(i.e. to diminish the fire area and property damage) and to stop secondary fires in lining
materials. Experiences from Japan show that sprinklers are effective in cooling down the area
round the fire, so that fire fighting can be more effective (refer chapter VI.3.4.3). It stated that
the use of sprinklers could be problematic because:
water can cause explosion in petrol and other chemical substances if not combined with
appropriate additives;
there is a risk that the fire is extinguished but flammable gases are still produced and may cause
an explosion;
vaporised steam can hurt people;
the efficiency is low for fires inside vehicles;
the smoke layer is cooled down and de-stratified, so that it may cover the whole tunnel;
maintenance can be costly;
sprinklers are difficult to handle manually; and
visibility is reduced.
It stated that due to the above issues, sprinklers must not be used in the region of a fire before all
people have been evacuated. It concluded that sprinklers cannot be considered as equipment
useful to save lives and could only be used to protect the tunnel once evacuation was completed.
As a consequence, it stated that sprinklers were generally not considered as cost effective and
that they were not recommended in usual road tunnels.
2.1.2. 2007 Systems and Equipment for Fire and Smoke Control in Road Tunnels
The relevant section in the 2007 PIARC publication Systems and Equipment for Fire and Smoke
Control in Road Tunnels [31] is chapter 6.4, Automatic Fire Suppression. It repeated the issues
raised in the 1999 publication and restated that if sprinklers are installed in a tunnel, they must
not be activated until all the people have evacuated. It did, however, note that experience in
Australia, the Netherlands, and Japan showed that sprinklers may be effective in cooling down
the area around the fire, so that fire fighting can be more effective and the risk of the fire
spreading to other vehicles can be reduced. It acknowledged Water Mist Systems as an emerging
technology and a variant to Deluge Systems, and recognised that new research work and technical
development were in progress at various locations around the world.
This section concluded that, at the time of documentation, an owner/operator who wants to
install new detection and new fire fighting measures must verify that they contribute to the
overall safety, and are compatible with the safety concept for that specific tunnel, as well as
ensuring the effectiveness of the proposed measures.
2.1.3. 2008 Road Tunnels: An Assessment of Fixed Fire Fighting Systems
The PIARC 2008 publication Road Tunnels: An Assessment of Fixed Fire Fighting Systems [30]
was intended to provide up to date information on the use of FFFS in road tunnels as current at
the time of documentation. It noted that significant research work had been conducted on FFFS
which provided a better insight into the advantages and disadvantages of these systems, and that
the goals of these systems were to slow down fire development, and to reduce or completely
prevent a fire from spreading from one vehicle to another, with the aim of improving conditions
8
for escape and rescue. It also recognised that as an outcome of the loss of life and extensive
damage to infrastructure from the then recent tunnel fires (i.e. Mont Blanc Tunnel - France/Italy,
24 March 1999; the Tauern Tunnel - Austria, 29 May 1999; the St. Gotthard Tunnel - Switzerland,
24 October 2001; and the Frjus Tunnel - France/Italy, 4 June 2005) that the need for further
improvements to tunnel fire management had been renewed. Post incident analysis of these fires
revealed common themes and that there were numerous opportunities for improving tunnel
safety following a fire incident. As a consequence, it was stated that FFFS may in some
circumstances be warranted provided a systematic approach, similar to that used for all tunnel
safety system components, was undertaken. Further, it was noted that before FFFS were installed
in a tunnel, a systems engineering approach to their integration into the tunnel safety concept
must be concluded. The interdependence between FFFS and other safety components must be
understood to ensure proper performance of the integrated tunnel safety system.
The report concluded that FFFS are one of many system types available to improve user safety
and infrastructure protection, and that the appropriateness of FFFS versus other system types
should always be considered. Installation of FFFS should only proceed where there is an effective
method, or methods, for detecting and precisely locating the fire. Without such capability, FFFS
will not work effectively. The review of determining if FFFS are appropriate should include
verification and validation of actual effectiveness, reliability, and performance, to ensure that
appropriate and informed decisions are made. The report recommended that the cost component
include installation and maintenance costs when undertaking a cost analysis.
2.2. OTHER REPORTS AND GUIDANCE
2.2.1. NFPA 502 - Standard for Road Tunnels, Bridges, and Other Limited Access Highways,
2014 Edition
Appendix E of NFPA 502 [20] Water-Based Fixed Fire-Fighting Systems in Road Tunnels
discusses FFFS.
Appendix E notes that it is now widely acknowledged that FFFS are highly regarded by
professional fire fighters and can be effective in controlling a tunnel fire by limiting the fire
spread. This is achieved by cooling down vehicles, to stop the fire from spreading to other
vehicles (i.e. to diminish the fire area and property damage), and to stop secondary fires in
tunnel lining materials. It also notes that the inclusion of water-based fire fighting systems can
act as a valuable component of the overall fire life safety systems in a tunnel.
It provides examples of where FFFS have been adopted in road tunnels and recommends that it
be considered as part of a package of fire life safety measures in long or busy tunnels where
an engineering analysis demonstrates that an acceptable level of safety can be achieved. Where
FFFS are applied, NFPA 502 recommends that they be activated within three minutes to prevent
development of a major fire. It also notes that FFFS should be part of an integrated approach to
the management of safety.
2.2.2. Other Relevant Reports
Other significant reports influencing the subject of FFFS in road tunnels include the UPTUN
Report D251 Engineering Guidance for Water Based Fire Fighting Systems for the Protection of
9
Tunnels and Sub Surface Facilities [26], SOLIT2 (Safety of Life in Tunnels) Engineering
Guidance for a Comprehensive Evaluation of Tunnels with Fixed Fire Fighting Systems [24],
and the BASt reports BASt 2014a [35] and BASt 2014b [36].
3. DECISION FACTORS
3.1. INTRODUCTION
When deciding whether or not to install any type of FFFS, the following must be examined:
compliance with local regulations and guidelines, including legal considerations;

global guidelines and safety standards;
the functions and roles of FFFS in the safety concept;
life safety;
asset protection and the protection required to assure the availability of the transport link;
flexibility for additional traffic regimes such as Dangerous Goods Vehicles (DGVs);
fire-fighting response;
the ability to adequately operate and maintain the system, including the roles, positions, and
responsibilities of the stakeholders;
the installation capital cost and/or life cycle cost, as well as the cost benefit from installing FFFS;
system reliability and redundancy; and
sustainability, as this may also be a factor in the decision.
Design considerations are addressed in chapter 4, page17.
3.2. COMPLIANCE WITH REGULATIONS, GUIDELINES AND LEGAL

CONSIDERATIONS
The tunnel regulations of most countries do not prescribe installing FFFS. In fact, the attitude of
various countries around the world to FFFS varies. Some countries choose to make it mandatory
to install FFFS in all road tunnels. Notably, this is the case in Australia (refer Appendix 1,
page44 for examples) and for longer tunnels in Japan (over 3 km). Equally, some countries
choose not to consider FFFS, and in some countries, the option of installing FFFS is considered
on a case by case basis by regulators and project stakeholders. This possibility may be expressed
in the relevant national standard or guidance, or may be treated by exception for each case.
Where FFFS are not mandatory and are permitted, the risk profile associated with the road
tunnels should be considered. Key factors include the location and characteristics of the tunnels,
the nature of the vehicle fleet, traffic volumes, and goods transported. Funding priorities also
vary from country to country and have to reflect the safety concerns, public attitudes to safety,
and policies specific to each country. It is relevant to note that the European Directive on
minimum safety requirements for road tunnels [38], which applies to 28 countries with widely
varying tunnel risk profiles, does not mention FFFS.
3.3. RISK ASSESSMENT
Each country uses differing means to determine how safe a tunnel is and whether that level of
safety meets its unique requirements and circumstances. This means that the level of safety
10
required, and the levels of safety achieved, vary from country to country and as knowledge and
circumstances change.
The installation and operation of FFFS does not make a tunnel safe, nor does not using FFFS
make a tunnel unsafe. A useful way to evaluate the level of safety achieved in a tunnel is to
conduct a risk assessment.
In most countries installation of FFFS is not prescribed by tunnel regulations. Therefore, they
typically are an additional safety measure which may be applied to mitigate specific risks or to
compensate shortcomings in tunnel design and other equipment. In the context of an integrated
tunnel safety design risk assessment, tools may be applied as support for decision making, to
demonstrate how and to what extent FFFS are suited to compensate specific risks, taking the
interaction with other tunnel safety facilities and safety relevant operational rules into account.
For this purpose, it is preferable that quantitative system based methods should be applied
containing simulation tools capable of modelling tunnel fires, smoke propagation and the effects
of FFFS on fire development, heat and smoke. This is discussed in the PIARC publication Risk
Evaluation, Current Practice for Risk Evaluation for Road Tunnels [37]. Such an approach can
also be applied to identify opportunities to attenuate other safety measures with overlapping
effects and demonstrate the improved efficiency. For example, some reduction in the overall
capacity of the ventilation system and the structural fire protection may be justified without
reduction of the overall fire life safety goals.
As noted in the publication on Risk evaluation for road tunnels [37], risk analysis is a tool for
assisting with decision making and has limitations related to simplification of data, assumptions
and risk weightings.
3.4. LIFE SAFETY
Life safety should be viewed from the perspective of the following groups:
tunnel users;
tunnel operations and maintenance staff;
emergency Services personnel; and
people external to the tunnel (e.g. people in buildings above the tunnel, or near the portal).
FFFS may improve conditions during a fire event by reducing the fire growth, the gas temperatures
and radiant heat fluxes. By reducing the fire growth and ultimately the fire size in megawatts,
the production of toxic gases, as well as smoke, may be reduced. This is particularly relevant for
self-rescue by tunnel users, however, to benefit self-rescue, early activation of FFFS is essential
to assure that the fire size remains as small as possible.
Tunnels should have an acceptable level of safety for all tunnel occupants. During a fire incident,
survivability in the fire zone is dependent on occupants escaping to a place of safety or reducing
and controlling the fire to an extent that it does not pose a hazard. In the area close to a fire, there
may be untenable conditions due to the effects of the fire (definitions of tenability are available
in the PIARC publication Fire and Smoke Control in Road Tunnels [29] chapter 1.5; and in
NFPA 502 2014 [20] appendix B). Untenable conditions for human life comprise heat (including
radiant heat) and oxygen content (or its replacement by CO2) and toxic gasses produced by
11
combustion, such as HCN and CO. Note that visibility is sometimes included as a tenability
criterion as it can affect wayfinding, however, if anyone is injured and unable to self evacuate
from the fire zone, this criterion is of less relevance than other criteria.
Untenable conditions close to a fire zone can occur regardless of whether FFFS are used or not.
The effects of FFFS on the production of toxic chemicals such as CO or more smoke production
are not clearly understood. Negative influences on these variables are counteracted by early
activation of FFFS. Furthermore, until a fire is totally extinguished, it may not be possible to
determine if the fire will flare up, or if some form of structural collapse may occur. In either
case, if there are any injured occupants in the fire zone, it is imperative to get them to a place of
safety so that proper treatment and triage can be accomplished. Anyone responding to a fire
event in a tunnel must first ensure their own safety before attempting rescue. The advantage of
FFFS is that they retard the fire growth rate and therefore the survivability of any injured
occupants is enhanced, as is the potential for responders to rescue anyone that is injured and
cannot self evacuate.
Tests and computational analysis demonstrate that the vertical impulse of activated FFFS may
transfer the higher concentrations of toxic gases in the upper levels of the tunnel cross section
towards lower parts; this effect has been observed in measurements of full scale tests [70].
Although no adverse effect of this phenomenon has been observed in real events involving
activation of FFFS, it again highlights the importance of the activation strategy and the ventilation
system for protection of life safety.
Visibility is also important for self-rescue. The application of FFFS will have an impact on
visibility through the presence of water where visibility is reduced. Another aspect is the impact
of FFFS on smoke. Smoke stratification, especially in the activated suppression zone, will be
degraded if not destroyed by employing FFFS. This may impede wayfinding and increase smoke
concentrations in the downstream zone. However, this has to be viewed from the consideration
that FFFS should be activated as soon as possible and that consequently, the temperatures, heat
release rates (HRRs), and volume of smoke produced are reduced, which all have a positive
impact on the tenability conditions in the tunnel.
On the other hand, there is at least a theoretical risk that activating FFFS could discourage
people evacuating from their vehicles due to their apprehension about the water discharge,
especially in situations when the potential threat of a developing fire is not yet perceived. In
documented events this risk has not been observed - but it does appear credible.
People must be persuaded to leave their cars for their own safety. Measures to improve this
response can be sound beacons, public address (PA) systems, guidance lighting, contour lighting
around the doors, strobe lights, and radio re-broadcast (RRB) and other similar systems [68].
The 2007 Burnley Tunnel fire (Melbourne, Australia) demonstrated that people can be persuaded
to leave their vehicles by use of appropriate cues [40].
If FFFS are used, it may be possible to optimise other fire life safety measures after detailed
analysis without compromising the safety requirements. FFFS are active systems that must be
maintained and operated to a high standard to work as intended. Preferably, the effects should be
addressed in a quantitative manner on the basis of a risk based approach. For example, in the
Northern Link Tunnel (Sweden), it was decided to maintain longitudinal ventilation and to install
12
FFFS to permit traffic congestion, and hence not to install a smoke extraction system. Another
example is the Felbertauern Tunnel (Austria), in which FFFS were installed as an alternative to
passive fire protection to protect the fresh-air duct that serves as an egress route. In addition,
FFFS may be implemented to reduce the heat-release rate of the design fire, which would reduce
the required capacity of the tunnel ventilation system. The application or availability of full-scale
fire test data in these instances may be relevant to have a reliable basis for such an approach.
If FFFS are used as a compensatory measure, their reliability, availability and maintainability
(RAMs) need to be accounted for when assessing their benefit.
Further examples of where fire life safety was mentioned to be a factor in installing FFFS are
provided in Appendix 1.
3.5. ASSET PROTECTION
With regards asset protection, the objective is to minimise the impact of the fire to the tunnel
structure and/or safety and services equipment. Protection of the tunnel structure aims at
reducing the damage from a tunnel fire event. In the worst case events, this could be preventing
a tunnel collapse with more disastrous effects as a consequence. This can be crucial for
underwater tunnels, or where large buildings or critical infrastructure are located over the tunnel.
This was one of the reasons for installing FFFS at critical locations on the M30 Tunnels in
Madrid (Spain).
Asset protection can also be considered in terms of reducing the down-time of a tunnel subsequent
to a fire. This may be particularly important where a tunnel is a critical part of the road network
or a toll road. These benefits were demonstrated by the Heavy Goods Vehicle (HGV) fire in the
Burnley Tunnel Toll Road fire (Melbourne, Australia) in 2007. The severe incident involving
multiple HGVs and cars, with immediate deflagration, was limited to less than 20 MW by rapid
activation of the installed deluge type FFFS. Subsequent modelling suggested that without FFFS
activation the fire would have rapidly grown to more than 100 MW [40]. The FFFS limited
damage to the structure and services, and allowed the tunnel to be reopened within a few days.
Similarly, this was one of the reasons for the installation of FFFS on some of the privately owned
tunnels on the A86 ring road in Paris (France).
FFFS may be considered as a compensatory measure in fire-engineering design. It is recognised

that passive measures are normally considered to be the most reliable, nevertheless, in some
circumstances, it may be possible to reduce the level of passive fire protection. This can be
useful, for example for refurbishment projects, where there are physical constraints such as a
lack of space, or operational constraints on when a tunnel can be closed for refurbishment works.
In some tunnels, FFFS have been used in the smoke-extraction duct to cool the exhaust air so
that the smoke extraction fans are not exposed to the high temperatures that otherwise might
render the fans inoperable.
Where FFFS are critical for asset protection, careful consideration must be given to their
reliability, and appropriate redundancy may be required to achieve the availability criteria
desired. Certainly, the consequences of failure of FFFS to operate on demand must be considered.
Passive fire protection may offer some advantages in this respect, but would not protect equipment
located within the tunnel space.
13
Some examples of where protection of the asset and transport link was mentioned to be a factor
in installing FFFS are given in appendix 1, page44.
3.6. ADDITIONAL TRAFFIC REGIMES
The impact of HGV fires can be reduced using FFFS and therefore can lead to a tunnel allowing
the passage of dangerous goods transport where otherwise this would not be permitted. However,
some aspects of HGV fires, while improved, may not be eliminated, such as the release of toxic
gases. The effectiveness of FFFS may not prevent extreme events such as a Boiling Liquid
Expanding Vapour Explosion (BLEVE), but can reduce the probability of its occurrence [12].
This was the main reason for installing FFFS in the Gnistngs Tunnel in Gothenburg (Sweden)
in which FFFS will also be engaged on fires with dangerous goods vehicles.
FFFS can reduce fire growth and fire spread between vehicles, which is particularly important
in a congested traffic situation. Further examples of where this was mentioned to be a factor in
installing FFFS are given in appendix , page44. These examples include the Felbertauern Tunnel
in Austria and the Roermund Tunnel in the Netherlands.
3.7. FIRE-FIGHTING RESPONSE
A fire hose has a limited ability to project a water jet onto a fire in a tunnel environment due to
the constraint the tunnel roof poses in limiting the height of the water jet. As a consequence, fire
fighters need to be closer to a fire in a tunnel environment than they would otherwise be in an
ambient environment. This exposes fire fighters to greater risk of the effects from the fire due to
their proximity.
However, even with specialised equipment which permits fire fighters to withstand higher
radiant heat, there are limits to the gas temperatures, levels of incident radiated heat flux, and
exposure duration that may be endured. This means that fire fighters may be unable to approach
a fire to reduce its size or even significantly reduce the radiated energy levels. In a Swedish
research project, it was concluded that it becomes difficult to fight fires above about 25 MW [14].
To fight large fires without putting fire fighters at risk, FFFS can be installed.
Although fire fighters may reach the tunnel quickly, commencement of fire fighting operations
may take a significant period of time in some tunnels. This time can be between 5 to 25 minutes
after arrival depending on the nature of the incident. Consequently, if the fire is allowed to grow
freely, it may quickly become too severe for fire fighter intervention.
FFFS will, in most circumstances, retard fire growth. The final extinction of the fire requires
intervention by the fire brigade. Procedures for fire fighter response should include the provision
that FFFS should not be turned off until it has been ensured that the fire is under effective
control.
In some cases, such as where the local fire brigade are located very close to the tunnel, the
additional benefits provided by the installation of FFFS may be reduced.
14
3.8. OPERATIONS, MAINTENANCE AND REGULAR TESTING
Maintenance of FFFS includes all technical, operational and administrative requirements to

ensure the system performs as specified during its design life. Regular maintenance of FFFS
commences immediately after the initial testing, commissioning and proving of the system.
Installation of FFFS can only be considered if the Tunnel Owner can commit to the provision of
supporting operation and maintenance regimes. To provide effective protection in the event of a
fire, FFFS must be adequately maintained, regularly tested throughout the life of the system, and
operated by suitably trained personnel. If this cannot be assured, then the benefits of FFFS will
be considerably diminished in a fire event, if not negated altogether. A strict routine of staff
training, system testing and maintenance of the system must be employed, or these activities can
be outsourced to a specialist company. Some testing of FFFS may also require some form of
tunnel closure (appendix 3, page68).
The decision when to activate the system is part of the overall operational strategy. Early
activation offers the greatest chance for fire control and can be important to ensure that FFFS are
not disrupted by the fire prior to their activation. Consideration should be given to methods for
confirming the fire location, and stopping traffic before system activation. To achieve this, clear
plans and procedures are necessary for operations.
Installation of FFFS should not adversely affect safety during normal operations. The
consequences of false activation should therefore be considered, as should system safeguards to
prevent this type of event. The importance of preventing false activation is clearly demonstrated
in the Central Artery North Area (CANA) Tunnel where the Deluge System was taken out of
service due to several accidental activations.
Further details on testing and maintenance are given in appendix 3, page68.
3.9. COST BENEFIT CONSIDERATIONS
The design life of a system can be defined as the period from the start of operational use until
the time the system needs total replacement. It should be noted that some components of the
system may need to be replaced a number of times during the design life of the total system. For
FFFS, the design life is typically 20 to 30 years depending on the materials used. Details on costs
in general are found in chapter 5.5, and examples from various tunnels in appendix 1, page44.
The decision whether or not to install FFFS may be supported by an assessment of the costs and
benefits. For this purpose, various approaches may be adopted. For example, decisions may be
based on the subjective opinions of key stakeholders, or an approach may be taken where the
costs and benefits are measured in some qualitative or semi-quantitative way, or a fully quantified
business case may be prepared utilising cost benefit analysis techniques.
Considering the subjective approach, a key stakeholder may advocate the installation of FFFS on
the basis that the system will limit fire spread, and reduce the potential fire and smoke hazards
in the tunnel. In some cases, the stakeholders perception of the safety benefits and public opinion
may be sufficient to determine the outcome.
15
At the other extreme, a formal business case might be needed that compares the capital, operating
and maintenance costs with the benefits of reduced injuries, damage and traffic disruption, over
a specified period such as the design life of the FFFS adopted or the tunnel. The output would be
an estimate of the overall cost benefit ratio.
The life safety benefits could be quantified on the basis of how many casualties would be
prevented, and valuations of the benefits of preventing casualties, as published by some national
authorities. These effects on risk (prevented casualties per year) could be quantified on the basis
of a quantitative risk analysis. The benefits for asset protection and transport route availability
could be quantified on the basis of major fires avoided, and the consequent repair and traffic
disruption costs avoided. The estimated costs of repairs would include the costs of repairing the
tunnel structures, renewing cables and replacing equipment, cleaning and re-painting, repairing
the road surface, plus the various associated costs of undertaking site works.
The cost of traffic disruption could be estimated by the amount of time lost for the traffic
participants, because of the non-availability of the tunnel that forces them to take an alternative
route. This can be quantified by the Value Of Time (VOT) for the traffic participants. The VOT
depends on the purpose of the journey being hindered. The VOT for recreational traffic is lower
than the VOT for business traffic and HVG transports. In countries like the Netherlands, Austria
and Switzerland, standardised VOT values (as well as standardised values for casualties and
wounded people in traffic/tunnel accidents and calamities) are used to allow for a standardised
approach for cost benefit analyses. A possible concern with a quantitative approach is whether
and how to take account of subjective issues such as public opinions about tunnel safety and the
availability of a transport route.
Since the fire risks and the benefits of FFFS will persist through the lifetime of the systems, it
would be appropriate to consider the whole of life costs of the system, which will include both
capital, operating and maintenance costs. The costs of periodic refurbishment or replacement
may also need to be included, depending on the timescales adopted for decision making purposes.
Certain factors would be expected to change over the period of interest such as the volume of
traffic flows, equipment and labour costs, and traffic diversion costs. Trade-offs with other fire
safety measures may be relevant
How each country conducts cost benefit analyses varies and reflects the large range of value
systems adopted by different cultures and authorities internationally. The important element is
that a cost benefit analysis be conducted.
Some authorities use standard procedures for decisions concerning highway safety, and may
choose to adopt the same practices when making decisions about FFFS, on the grounds of
consistency.
It is possible that the presence of FFFS as part of the tunnels overall risk mitigation strategy can
reduce insurance rates for accidental losses and transport disruption in cases where such
insurance policies exist. This is the case in Australia.
16
3.10. SUSTAINABILITY
Sustainability involves broad concerns about economic, social and environmental objectives.
For transport infrastructure, construction and operation increasingly include measures to
improve sustainability by reducing environmental impact over the operational lifecycle. Such
impacts are typically assessed in terms of carbon emissions. Most efforts to date to assess carbon
emissions from facilities focus on normal operating conditions, but some consideration has been
given to the influence of fires and FFFS.
To assess the impact of FFFS, it is important to understand the overall context of carbon
emissions. The total carbon emission over the lifecycle of a road tunnel will include the sum of
emissions from:
construction (including materials, transportation and equipment usage);

normal operations and maintenance (notably power consumption);
periodic refurbishments (equipment and transportation for disposal of redundant materials, and
further new materials, transportation and equipment usage); and
tunnel fires (and other incidents), mitigation responses and subsequent repairs.
The direct benefit of tunnel FFFS reducing the fire risks and corresponding carbon emissions is
partly negated by the carbon emissions associated with FFFS themselves and hence considered
to be marginal. To achieve a larger impact on sustainability, installing FFFS would have to be
associated with other design optimisations such as reducing the design fire size, slimmer or
smaller structures (i.e. requiring less concrete or excavation volume), or reduced equipment costs
by a reduction in equipment size or capacity.
To quantify the impact of FFFS, it is therefore necessary to quantify the potential impact of fire
risks on the overall carbon emissions over the lifetime of the tunnel. A basic framework for such
calculations is illustrated in appendix 6, page79.
3.11. LEGAL CONSIDERATIONS
As is the case with all asset protection and life safety equipment, if injury or tunnel damage is
caused by fire, legal investigations will likely consider the circumstances of the installation of
FFFS (or the decision not to install), and any differences between how they were understood to
perform and how they actually performed (refer [27] and [28]).
In some countries or regions, where FFFS have not been installed, a legal investigation will also
likely consider whether FFFS should have been installed to prevent the loss in the first place.
Similarly, where loss occurs with FFFS installed, the circumstances of the loss. However, the
European Directive on minimum safety requirements for road tunnels [38] does not mention
FFFS and within this framework, it could be assumed that FFFS are not compulsory to be
considered. There are legal presumptions that FFFS, like all equipment, are optional, only
installed because they are needed to manage fire risk, fit for the purpose, and that they will be
integrated, operated and maintained correctly. Depending on the legal environment, Tunnel
Owners and Operators may be advised to ensure that they articulate clearly why FFFS have been
installed or not installed, and how FFFS are to be tested, maintained and operated.
17
Prior to a decision about FFFS being installed, documentation should record the factual
circumstances of the process by which the installation of FFFS was considered and agreed. For
example, if FFFS are installed to protect a tunnels tolling revenue and for instance not needed
for safety reasons. If there is no need for FFFS because the level of risk is assessed as not
warranting FFFS, then the documentation should record that fact. If FFFS are installed to protect
vulnerable structural portions of the tunnel, it should be ensured that this is recorded in the
project documentation accordingly. If FFFS form an integral part of achieving the requisite level
of safety in a tunnel, this fact should be recorded clearly.
Furthermore, the range of FFFS and the range of opportunities to install and operate them means
that the anticipated performance of such systems varies and should be documented accordingly,
along with the anticipated operational regime to achieve that performance. Documenting both
the strengths and weaknesses of FFFS and their contribution to achieving asset protection and
life safety ensures that it is seen as but one of a range of integrated risk mitigation strategies
which neither individually nor collectively provides absolute safety.
Where FFFS are used as part of an integrated approach to manage safety and or asset protection,
it is likely that their installation and use will impact other aspects of the tunnel design and
operational strategy. For example, FFFS may reduce the size of the design fire, or the rate of fire
growth used in the civil works component of the design package. Such impacts must be recorded
and documented to ensure that during the operation of the tunnel, the interdependence of
operable FFFS and factors such as structural integrity and evacuation performance is remembered.
Known weaknesses of the technology of FFFS should also be documented and put in context.
For example, there are a range of credible false activation and other scenarios which could be
identified.
Any legal investigations into FFFS installation, integration, false activation, intentional operation,
expected performance or maintenance will be heavily influenced by documentation which
pre-dates an incident. Like other safety systems in tunnels, it is important to record and evidence
the engineering rationale that relates to the selection, installation, integration, maintenance,
testing and use of FFFS.
4. DESIGN CONSIDERATIONS
This section is written from the perspective that the decision to install FFFS has already been
made. Risk assessments are useful in informing the designer regarding design parameters as
discussed in chapter 4.1, page18 but they can also be used to provide a basis for determining if
FFFS provide benefit to a tunnel system. The information contained in this section is intended
to provide the decision maker an understanding of key design considerations including their
influences on other tunnel systems.
Once it has been determined to install FFFS, the designer must establish a viable working design,
and consider the following design issues in the process of developing the design of the FFFS:
design fire;
risk assessment;
type of system;
18
water suppression characteristics (mist or deluge);

water supply including possible hydrant and/or standpipe systems;
drainage;
space considerations;
fire detection/activation strategy;
environment;
system integration;
interaction of FFFS with ventilation; and
other factors.
The objectives of FFFS may fall within a range; for example from extinguishment at one end of
the scale to modest cooling at the other end of the scale, depending on the specific case. These
requirements may be categorised (similarly to the categorisation in NFPA 502 [20]) as:
fire suppression;
fire control; and
thermal exposure reduction.
The term suppression may be broadly defined as a sustained reduction in the HRR following
activation, control may be defined as a reduction of the HRR such that the peak is less than
would be expected without FFFS, and thermal exposure reduction is the cooling of gas and
shielding of radiation to reduce the exposure of people, structure and/or equipment without
affecting the HRR. NFPA 502 [20] uses similar definitions, with thermal exposure reduction
split further to volume cooling and surface cooling.
The design fire criteria will be derived to support the qualitative objectives, and will depend on
the tunnel and traffic characteristics. Criteria should include aspects such as:
design fire HRR and fire growth characteristics;

fire location;
ventilation conditions;
time from ignition to the commencement of suppression at the fire site;
the fire HRR at the commencement of suppression;
FFFS operating principles; and
fire test protocol, including fire location.
The qualitative objectives and derived design criteria may then be developed and defined
quantitatively in the development of the design to, for example, establish the required water
application rates, detection performance and activation strategy.
4.1. RISK ASSESSMENT
The designer should consider the findings from chapter 3, Decision Factors, page9 with regards
Risk Assessment and consider implications to the design approach. If necessary, additional risk
analysis may be required.
19
4.2. TYPE OF SYSTEM
There are two main types of water based FFFS installed in road tunnels today. These are Deluge
Systems and Water Mist Systems. The main mechanisms of fire suppression for these two types
of systems are different (chapters 1.4.1 and 1.4.2, page4). Where the fuel is in liquid form, FFFS
may increase the surface area of the fire.
The type of system drives fundamental design values such as water application rate, which in
turn defines the required water supply in addition to the capacity of the drainage system. Deluge
Systems require a higher total flow rate and hence may require additional water supplies and
may have a more significant impact on the drainage systems. Water Mist Systems require a
lower water supply rate, but the pressure requirements can be higher.
The choice of system can be affected by space, drainage and water supply issues, as well as the
suppressive capability of either type of system. Where liquid spills occur, drainage systems can
reduce the pool surface size which can reduce the fire size. In retrofit applications particularly,
the drainage and water supply capacities can be crucial, and if limited, would tend toward the
choice of Water Mist Systems. Power supply is also an important factor since higher pressure
Water Mist Systems require significant pumping capacities in comparison with Deluge Systems
(i.e. more energy is required to both pump the water to the nozzle and convert the water to mist
by high pressures, than to pump water to an open nozzle).
Illustration 1 Deluge electric pump and panel

Marina Coastal Expressway, Singapore
The risk assessment process may be used to establish the type of FFFS that will be installed.
Further information on types of FFFS can be found in appendix 2, page58.
4.3. WATER DISCHARGE CHARACTERISTICS (MIST OR DELUGE)
Water discharge characteristics are defined by water droplet size, pressures, piping layout, nozzle
design and water application rate. There is a significant difference in these characteristics
between Water Mist Systems and Deluge Systems.
20
There is no single deterministic method for determining the water application rate. Typically the
water application rate is a decision combining current practice, full scale testing and/or
engineering analysis. The water application rate typically depends on the objectives for the
FFFS. In some countries, such as Japan, the water application rate is pre-defined and varies from
2.5 mm/min to 12 mm/min. For further details see appendix 2, page58.
Illustration 2 Deluge discharge

4.4. WATER SUPPLY
FFFS require a reliable water source and an effective distribution system. This generally requires
suitable pressure and flow from the town mains, or some form of water storage. Where water from
the town mains has sufficient flow but insufficient pressure, booster pumps will be required. Where
water storage is required, either to supplement the town mains or to provide the required volume of
water, sufficient capacity must be provided for the required duty. In this case, booster pumps will
generally be required unless the hydrostatic head is sufficient to provide the pressure (e.g. where the
water storage tanks are above the tunnel, or in an elevated location on a hill). The amount of water
available may influence the type of FFFS selected. Tunnels can have FFFS as well as hydrants and/
or standpipes. When this occurs, the designer needs to address the additional water flow and pressure
requirements of all systems. FFFS may also require independent water supplies for higher reliability.
In addition to the required water supply, the designer needs to ensure that a sufficient volume of
water is available to service all system requirements. The volume of water required may be specified
(generally a minimum of one hour) or determined by analysis of the appropriate design fire HRR
and response. This may require the provision of water tanks to provide sufficient volume of water.
21
Illustration 3 Deluge water storage tank

4.5. DRAINAGE
With the application of FFFS, additional fluids will be added to the tunnel environment. These
must be carried away efficiently and effectively. There are a number of issues to consider in
designing an adequate drainage system. In addition to any normal design requirements, the
capacity of the drainage system should be designed for the total water flow from the FFFS plus
additional water sources such as hydrants that may operate simultaneously, and importantly,
fluids that may be part of a flammable or combustible liquid spill. A properly designed drainage
system can capture this liquid spill, reducing the pool surface size and therefore reduce the burn
time and final fire HRR. Importantly, the drainage system should incorporate some measure by
which fire is not propagated through the drainage system to some other location. For existing
tunnels this provides a limitation to the design, but the risk of limited flooding may be acceptable.
4.6. SPACE CONSIDERATIONS
Space outside of the traffic envelope must be allocated for water mains, distribution piping,
valves and nozzles. Height clearance for overhead installations must be such that overheight
vehicles would not strike the pipework. Valve cabinets must be provided which can take
significant space in the tunnel cross-section. In some instances, valve cabinets can be installed
in cross-passages. The space requirements can drive the cross-passage size.
These issues can be more challenging in existing tunnels. This can potentially influence the type
of FFFS that are selected. In high pressure Water Mist Systems, the pipework might be smaller
in diameter, making the installation easier. Availability of ventilation ducts to route pipework
may also be a factor in the choice of system, however, if this route is adopted, any effect on the
ventilation system performance must be considered.
The allocation of space to FFFS must not interfere with the function of other systems in the
tunnel and also must not create an additional hazard for vehicles in any way. For instance, the
pipe network must not interfere with the function of jet fans, CCTV or incident detection, nor
must the routing of pipework for FFFS around other systems adversely affect the hydraulic
efficiency of the system.
22
4.7. FIRE DETECTION AND ACTIVATION STRATEGY
A means of fire detection is essential to facilitate early and accurate activation of FFFS at the
correct zone. Fire detection methods vary from tunnel to tunnel and can be automatic, operating
remotely and providing an alarm without human intervention, or manual, where human
intervention is required to initiate an alarm. Types of automatic detection systems include linear
heat detection, spot heat detectors, smoke detectors, and CCTV based imaging technology for
detection of flame or smoke from a fire. Manual fire detection systems include fire pull boxes
and operator manned CCTV. In some cases, for both automatic and manual detection systems, a
combination of a number of different techniques is used to minimise false alarms. However, care
should be taken if this strategy is adopted. Different detection techniques have different detection
times and therefore there is a possibility that a fire may be allowed to develop while waiting for
a second alarm to verify and confirm the event.
Activation of FFFS can be:
automatic, where on receipt of an alarm from a detection system the FFFS will activate
automatically without delay;
semi-Automatic, where on receipt of an alarm from a detection system the FFFS will activate
automatically, but where a time delay is given for potential manual intervention either to
activate the FFFS or to cancel the alarm prior to automatic activation; or
manual, where activation of FFFS requires human intervention for the system to function.
Therefore, regardless of the detection methodology, there are a number of decisions that are
required concerning the activation sequence and the mode of FFFS activation. All the methods
of activation described are used in different parts of the world.
False activation of FFFS should be avoided as this may cause a collision. However, during an
incident, where FFFS are activated as soon as possible, the incident vehicle should have stopped
before the FFFS are activated and therefore most other vehicles should either be stopped or
travelling at reduced speeds.
A fire detection system should be reliable and at a minimum identify the location of the fire such
that the correct zones of the FFFS can be activated. Where the fire zone is not correctly identified,
the tunnel operator may be forced to correct the activated zone. It is therefore important that the zone
can be able to be activated and deactivated remotely. CCTV can be a useful tool in this process.
Current practice uses open style FFFS and not thermally operated elements. The use of thermally
operated elements in a tunnel environment could result in the activation of nozzles that are not
directly over the fire due to the longitudinal air velocities activating thermally operated elements
downstream of the fire event. Even if thermally operated elements are activated over the fire
zone, it is possible that the designed water suppression volume is not delivered to the fire site as
more nozzles may be activated than permitted in the design and consequently, insufficient water
is available for the design fire HRR. With thermally operated elements, this cannot be corrected
as there is no way to deactivate individual nozzles remotely.
23
4.8. TUNNEL ENVIRONMENT
In areas prone to freezing temperatures special arrangements may have to be made. There are
multiple methods available for preventing freezing of pipes. These include trace heating,
anti-freeze additives, circulation pumps with or without water heating, and insulation. A dry
system might be chosen to avoid freezing issues, however, this may impact on the time required
to provide water to the fire zone. In addition, the potential and effect of high temperatures also
needs considering for the pump room, water supply, the water main, and the valves.
In many countries, water effluent discharge to local natural resources or even a local municipal
sewer is sensitive. These issues must be addressed in the design of the system and appropriate
measures incorporated to capture waste streams so that they can be processed appropriately.
Though not addressed within this report, additives to water also have design implications, such as
increasing the potential for corrosion. Microbiologically induced corrosion may be possible in some
areas. Use of antifreeze, if considered, should meet acceptable standards for antifreeze concentration.
Corrosion is an issue in all road tunnel environments. This effect is exacerbated near salt water
bodies, and in areas where salt is used for de-icing. In some countries such as Germany, all
tunnel hardware is required to be a high grade stainless steel. Each road tunnel is unique and the
circumstances surrounding the tunnel should be considered with respect to material selection
and other corrosion protection measures.
The application of FFFS in sub-zero temperatures needs consideration for drainage where pipes
and drain covers may freeze, and for road surfaces where icing may result in additional efforts,
such as de-icing, to minimize possibly dangerous road surfaces.
4.9. SYSTEM INTEGRATION
FFFS must work in harmony with the other tunnel fire life safety systems. It is essential that the
interdependencies between FFFS and ventilation systems be properly understood and control
systems integrated for effective use. These systems include fire detection, ventilation, CCTV,
traffic management and other communication systems. FFFS should not be seen as a discrete
independent risk mitigation strategy but should be seen as an integral part of the fire fighting
systems respective safety system. A discussion of these factors is given in table 1.
TABLE 1 - FACTORS ASSOCIATED WITH SYSTEM INTEGRATION

Factors Comment
Traffic control systems, and operational procedures, must be integrated to be compatible with FFFS
Traffic Control
activation so that activation only occurs with stopped traffic.
If FFFS is adopted, the ventilation concept and strategy should be reviewed, and may need
Ventilation
modification. Additionally, the system capacity may be affected.
Power Supply Higher pressure systems may require additional power to drive the pumps
Visibility may be impaired by operation of FFFS and consideration should be given to upgrading the
CCTV
systems if this is deemed to be an issue.
Activation of system should be prompt and in the correct zones. Manual activation would involve the
Operation
operator in identifying the location, etc.
Detection Needs to support operation of system in terms of accurately identifying fire location.
Water Supply Water storage requirements may be considerable and this issue could be problematic in a retrofit.
Needs to provide sufficient drainage capacity and containment. Considerations include the removal of
Drainage surface water and spillages, the capacity of the drainage line, fluid containment or outfall, and
ensuring fire is not propagated through the drainage system.
24
Depending on a risk assessment and the anticipated performance of FFFS the design fire HRR may be
Design Fire
reduced in magnitude.
Tenability (visibility, temperature, toxicity and radiation) is affected by operation of FFFS. FFFS cause
Tenability
conditions close the fire zone to be impacted but improve conditions further away.
Connected with operational issues and influenced by tenability, user action or inaction, audible and
Egress
visual communication methods and wayfinding.
For activation of FFFS, the fire detection system must be able to detect a fire within a prescribed
length of tunnel, which drives the design of the fire detection system.
Since the location of the fire is integral to the response of multiple tunnel systems, ventilation
activation zones, zone of FFFS, detection zones, wayfinding zones and CCTV should all be
integrated for optimal performance.
4.10. INTERACTION OF FFFS WITH VENTILATION
The tenability of the tunnel environment depends on factors such as smoke movement, toxic
gases, gas temperatures, radiation and visibility. It is essential to consider the tenability
implications of the use of FFFS when designing or activating these systems. Research has
demonstrated that FFFS will reduce the air temperature and radiation from a fire but decrease
visibility by disrupting the stratified smoke layer. As a consequence, shorter visibility distances
may result and systems to improve the identification of egress doors, such as sounders or strobes,
may be required.
Rapid deployment of FFFS is recommended to ensure that fire growth and HRR is minimised.
This results in minimal impairment to visibility and provides maximum protection to the tunnel
infrastructure and can allow more flexibility in design by providing the required level of safety.
This approach is adopted in countries such as Japan and Australia.
Experimental measurements have shown that the HRR can be reduced by 50% or more [75] [76],
and the heat transported by convection can be reduced from, typically 70%, to a maximum of
50% [72] when FFFS are operated. This effect reduces the Critical Velocity required for the
control of smoke from a suppressed tunnel fire relative to a given design fire HRR.
When operating FFFS in a tunnel, allowance must be made for the fact that operation of FFFS creates
extra resistance to the airflow. However, proper application of FFFS should retard the fire growth rate
and hence the maximum fire HRR should be similarly controlled or reduced. This effect may
also be taken into account in designing the ventilation system, if only to note this safety factor and
flexibility in the design.
4.10.1. Ventilation Systems

In general, two types of ventilation systems are used in road tunnels:
longitudinal ventilation system; and

transverse ventilation system.
Longitudinal Ventilation
A longitudinal ventilation system creates an airflow within the tunnel to push the smoke and hot
gases, in one direction, away from the stopped vehicles and evacuating passengers (illustration4).
25
The goal of a longitudinal ventilation system is to create sufficient airflow to oppose the buoyancy
forces of the hot smoke and thereby prevent the smoke from backlayering in the direction opposite
to the ventilation system airflow. The velocity of air required to achieve this airflow within the
tunnel is known as the Critical Velocity. Activation of FFFS in the presence of a high air velocity
may cause the droplets from the FFFS to be displaced, with the extent of displacement a function
of air velocity, droplet size, and the momentum of the water droplets. However, fire tests and
research projects have shown that this is not an overly serious phenomenon due to the downward
velocity of the water droplets. It is also largely mitigated by minimising the longitudinal air
velocity to that which is sufficient to ensure Critical Velocity (i.e. prevent backlayering).
This prevents over design of the ventilation system and is good practice in any event.
Illustration 4 - Smoke displacement with longitudinal ventilation

PIARC 2008R07 - Road Tunnels: An assessment of fixed fire fighting systems
Transverse Ventilation
Transverse ventilation systems aim to extract the smoke by openings usually located at the
ceiling or walls so that smoke is removed from the region directly at the fire, while forcing
airflow toward the fire on both sides. Illustration 5 illustrates the general principle of this type
of smoke control when steady conditions have been established. Initial air movement prior
to activation of the ventilation system may cause smoke to move further along the tunnel before
control is established. Additionally, for unidirectional traffic, the transverse ventilation system
may be operated to provide a longitudinal airflow to protect traffic stopped behind the fire event.
It should be noted that FFFS will disrupt the smoke layer (if any) and will affect the smoke
extraction system. On the other hand, FFFS will reduce the fire HRR and hence reduce the
smoke production.
Illustration 5 - Smoke extraction with transverse ventilation

PIARC 2008R07 - Road Tunnels: An assessment of fixed fire fighting systems
4.10.2. Time Scale Influence on Interaction of FFFS with Ventilation

The fire will grow for an undetermined time from ignition until detection. After detection, the
FFFS and ventilation system can be activated as well as other activities required by the Emergency
Management Plan. The ventilation system will take time to reach full flow capacity, however,
there will be an additional time period that must be considered for the ventilation system
to effectively commence smoke management. These time periods should be minimised.
Immediate activation of FFFS is desirable for control of smoke and to minimise the fire HRR,
however, some jurisdictions may require a delay. Designers must account for this timeline and
26
ensure the owner understands the effect of delay of activation of the FFFS (typically 60 to
90 seconds from the time the systems are activated to reach the full application rate).
During the self evacuation phase, and particularly before the first responder arrives to assist
those who have not self evacuated, the most important factor is the element of time and how the
critical factors noted below change or evolve as a function of time:
the progressive growth of the fire until it attains its maximum FHRR;
tenability conditions within the tunnel; and
the movement of occupants towards the escape routes.
Understanding these spatial and temporal changes is of critical importance in analysing the
interaction between FFFS and the tunnel ventilation system.
A classical method for representing the interaction between the critical elements listed above is
a tenability diagram (illustration 6, which shows tenability conditions relative to the fire site at
3,500m, plotted against time [39]). Tenability diagrams, used as part of risk assessment
methodologies, graphically demonstrate:
fire growth, including changing contaminant levels and temperatures;

tenability conditions, represented by visibility, temperature, CO levels;
the evacuation process; and
the ventilation strategy, represented by the movement of air.
The greatest difficulties associated with this approach are selecting the input parameters to be
adopted in the models or in the test, and avoiding an unwieldy or impractical matrix of scenarios.
If these difficulties can be overcome, then this type of tool has the potential to provide a better
understanding of the interaction between FFFS and a tunnel ventilation system.
Illustration 6 - Example of a tenability diagram

27
4.10.3. Interaction with Ventilation
It is difficult to give general criteria or recommendations regarding the interaction of FFFS with
a ventilation system. Specific risks must be evaluated on a project-specific basis using risk
analysis, a scenario-based approach or other techniques adapted to the particular characteristics
of the tunnel and including the ventilation system and alternatives to FFFS selected.
The most common objectives of FFFS during the evacuation phase are to limit the extent of the
fire, its rate of growth, and to maintain tenability for people who are evacuating, as well as
emergency service personnel who are entering the fire zone. In addition to controlling the spread
of the fire until emergency services arrive and commence fighting the fire, FFFS offer benefits
such as increasing the fire resistance of the infrastructure, limiting the extent of fire damage and
fire spread, and reducing the time necessary for repairs. However, inappropriate or poorly timed
activation of FFFS can result in undesirable consequences.
Appropriate design should be undertaken to ensure that the correct zone is activated for the fire,
and an engineering analysis should be performed to determine where FFFS should be installed
and when they should be activated to avoid undesirable consequences.
Some general observations can be made for a few key scenarios.
Unidirectional Tunnel with Longitudinal Ventilation and No Traffic Downstream of the Fire
Site
Without FFFS activation:
downstream conditions: Expected to become untenable.
upstream conditions: Expected to be safe unless the design fire HRR is exceeded.
With FFFS activation:

downstream conditions: May become untenable. The smoke is cooled due to activation of
FFFS and the smoke layer drops. Additionally, the air temperatures decrease relative to the
case without activation of the FFFS. FFFS also prevent fire spread from the point of origin.
upstream conditions: Safer than without FFFS. FFFS can assist the ventilation system by
reducing the final maximum fire HRR and by reducing heat radiation. FFFS also prevent fire
spread from the point of origin.
Bi-directional Tunnel, or a Unidirectional Tunnel with Longitudinal Ventilation and Traffic

Upstream and Downstream of the Fire Site
Without FFFS activation: Safe evacuation conditions may be unachievable except for a short
duration during the initial fire growth phase.
With FFFS activation: The FFFS can increase the level of safety by reducing the fire growth
rate, but can also have an adverse effect by destratifying the smoke layer. The outcome depends
on the location of people relative to the fire, distance to egress, the fire HRR, and smoke
development. FFFS also prevent fire spread from the point of origin.
Unidirectional or Bi-directional Tunnel with Transverse Smoke Extraction

Without FFFS activation: Safe evacuation conditions may be achievable for a short period
of time if the location of the fire is accurately identified and the ventilation strategy properly
selected. However, the amount of smoke generated by a fire may exceed the design extraction
28
rate of the ventilation system. In this case, the ventilation system alone may not be able to
handle the smoke.
With FFFS activation: As with the longitudinal ventilation system described above, FFFS
should reduce the fire size and have similar effects. As the smoke is cooler and at a higher
density, the transverse smoke extraction system needs to be specifically designed. FFFS also
prevent fire spread from the point of origin.
Summary
Regardless of the ventilation system type, if FFFS are properly designed, installed, well-integrated
with the other safety equipment in the tunnel, and managed in accordance with established
procedures, tunnel conditions during the evacuation phase can be stabilized and conditions
during the fire suppression phase can be improved by FFFS. Further information on tunnel
ventilation response can be found in the PIARC publication Road Tunnels: Operational Strategies
for Emergency Ventilation [32].
4.11. FURTHER ASPECTS
There are additional design considerations that need to be addressed. Since FFFS can be a
fire-life safety system, usually some redundancy of the system is provided. This can mean
backup pumps and redundant water supply, reliable control systems and additional power
requirements. Some form of reliability assessment should be undertaken as part of the design.
This may, for example, be a RAMs analysis. The functional requirements of the FFFS may affect
the RAMs analysis. For example the reliability requirements for FFFS used for life safety may
be greater than where the system is used for asset protection.
Even though FFFS are to be expected to be activated as soon as possible, it is still expected that
FFFS will continue to operate until the fire is suppressed or extinguished. Therefore, FFFS may
be expected to function in a very hot environment and material selection should be carefully
considered so that an elevated temperature environment can be endured while operations are not
hindered. Usually water flow through the pipes will help to mitigate high temperatures, but in
the event that the wrong zone is activated, the piping can be exposed to high temperatures.
Where the pipe can be exposed to high temperatures, materials such as PVC piping, rubber seals
for pipe connections, and other components may fail under high heat.
A maintainable system is essential to the operational resilience of the system. The FFFS should
be designed such that they are readily maintainable. Further information can be found in
Appendix 3, Maintenance and Testing, page68.
5. SYSTEM DEFINITION / PROCUREMENT
This chapter considers the performance and contractual issues involved in specifying and
procuring effective FFFS for installation in a road tunnel.
Uncertainties in conditions of contract can potentially lead to conflicting interpretations of the

project requirements and processes by the organisations involved. This can impact the design of a
system and its installation and testing. Ambiguity over the system definition and acceptance criteria
can lead to project delays, increased contractual claims, and disputes. From a safety perspective,
such ambiguities can result in the fundamental safety objectives not actually being achieved.
29
This chapter highlights key points that should be considered by the stakeholders and by clients
in particular, in order to overcome these problems.
5.1. SYSTEM DEFINITION
A first point to consider is whether particular FFFS should be specified or the choice left open
to bidders. A benefit of specifying a particular type of system is that the bids from different
contractors can be compared more readily. Alternatively, the question of which type of system
can be left open for bidders to propose alternative solutions, in which case the requirements will
need to be expressed in high level terms with performance specifications. This decision may be
influenced by the system complexity and the available maintenance providers. For whichever
type of system, the design should ensure compatibility of individual components before
modifications or upgrades are made, and regardless of the type of system, obtaining warranties
should be considered.
For procurement purposes, the system needs to be defined clearly in specifications. Useful
guidance to support the preparation of a specification includes the documents published in 2006
and 2007 by the European UPTUN research project [67] [26], and the SOLIT2 guidance [23]
published in 2012. These describe the general arrangements of FFFS, the interactions with other
systems, system design and maintenance concerns. Of course, such guidance will not necessarily
take into account the particular circumstances of a specific tunnel. Some key points to be
considered when preparing the specification are outlined in the following sub-sections.
5.1.1. Fully Engineered System

The installation of FFFS may be engineered in accordance with a standard such as NFPA 13 [19]
or NFPA 750 [21], in a similar way to the design of fire suppression systems for industrial
facilities in general. Such standards set out detailed design requirements including the nature
and adequacy of water supplies, selection of nozzles, fittings, piping, valves, and all materials
and accessories. This approach offers the benefit of minimising uncertainties in system definition.
However, the existing standards are not sufficient for the development of fully engineered
systems for road tunnels and may need supplementing for such applications.
Such standards do not take into account tunnel specific issues, such as the dynamic loading of
an installation due to traffic and ventilation, or atmosphere induced effects. These issues need to
be addressed by the specification.
Another potential issue is that the water application rates recommended in such standards are
derived for a range of industrial and building hazards, not for road vehicles. Shielding effects
tend to be a particular concern when dealing with vehicle fires. Without appropriate testing, or
existing fire testing results for similar circumstances, the actual fire fighting performance of a
fully engineered system is likely to be uncertain. Testing is required to provide assurance of fire
fighting effectiveness. In addition, modelling is now being used in some projects as part of the
assurance process (chapter 6.2, page35).
5.1.2. Off the Shelf Systems

A FFFS contractor may offer a system that is effectively off the shelf. Such a system may have
been fully engineered for another tunnel or facility, but will not have been tailored for the specific
tunnel. Consideration will therefore need to be given to the amount of additional design needed
30
to reflect the geometry and constraints of the specific road tunnel in question. This could be a
significant issue when retrofitting a system into an existing tunnel. Potential issues include
conflicts between the proposed systems piping and equipment, and the existing tunnel equipment.
For example, piping routes may be blocked by the presence of tunnel luminaires, cable trays, jet
fans and tunnel message signs. Piping may have to be diverted around existing equipment,
which may lead to nozzles being sited inappropriately and the hydraulic efficiency of the system
adversely affected.
As for fully engineered systems, the effectiveness of an off the shelf system will be uncertain
unless appropriate testing or modelling is undertaken, or fire test results are already available
for similar circumstances. For an existing tunnel, such testing should reflect the actual nozzle
arrangements.
A key issue for clients is that it may not be obvious at the tender stage that a system being offered
is really an off the shelf system with its inherent uncertainties. Of course, there may be commercial
benefits in terms of reduced capital cost.
5.1.3. Performance Specification

An alternative or supplementary approach to an engineered or off the shelf system is to specify
the actual fire fighting performance to be achieved by the FFFS. The performance is defined in
terms of selected parameters such as:
test material;
HRR versus time;
the time between commencement of the fire and activation of the FFFS;
tenability conditions, such as temperature, radiant heat flux, visibility and concentrations of
combustion products all typically evaluated at head height;
maximum temperatures at specified points (including equipment locations) on or within the
tunnel structures; and
mechanical, hydraulic and electrical requirements (for example as defined in SOLIT 2 [24] and
UPTUN guidance [26]).
The tenability and temperature performance criteria would be specified at particular times after
activation of the system, for example the maximum fire HRR and maximum temperatures after
activation of the system. Time-varying criteria may be specified if the objective is to suppress or
control the design size fire within a certain time period.
In the absence of existing fire test results for similar circumstances, the performance will need
to be demonstrated. SOLIT2 [24] offers guidance on test protocols for fire tests. The test protocol
needs to consider the tunnel geometry, the test fire load and ignition arrangements, and the
ventilation characteristics (including control protocol). The timing of system activation must be
consistent with the operating procedures and reflect the means of detection and the traffic
management systems, procedures and self-recue requirements. FFFS contractors may be
unfamiliar with such details and should be briefed to avoid any misunderstanding. It is important
that the immediate review of test results not be rushed. Adequate allowances for review and
repeat testing should be included. It is important to make sure that all stakeholders understand
the philosophy, the implications for performance in the real tunnel and the contractual situation.
31
A further key issue is that the acceptance criteria need to be agreed in advance. Ambiguity over
test results and acceptance criteria can lead to project delays, increased contractual claims, and
disputes.
5.1.4. Interfaces with Other Systems

Road tunnels typically incorporate a range of tunnel structures and mechanical, electrical and
traffic management systems, which are separate from, but may have interfaces with, FFFS.
These include:
tunnel structures;
automatic fire detection system (such as a linear heat detection system);
automatic incident detection systems (such as video or radar-based systems);
traffic monitoring systems;
traffic control systems (such as variable message signs, tunnel closure barriers, signs and
signals);
tunnel ventilation system;
water supplies; and
drainage systems (including sumps and disposal systems).
The performance specifications for these associated systems could be important factors to
consider when preparing the performance specifications for FFFS.
Problems are more likely to occur if the FFFS contractor lacks familiarity with the general
principles and practices of road tunnel design and operations. Frequent design reviews and
interface management meetings should be conducted throughout the design and checking/
approval phases.
5.1.5. Spatial Constraints

FFFS will require space for their installation, both inside and outside of the tunnel. Within the
tunnel bore, space will be needed for piping and possibly also for section valves. Outside of the
tunnel, space may be needed for pump house(s), vehicle access routes, valves, water tank(s) and
water supply connections, plus power supplies and other associated equipment. In addition,
provisions may be required for manual activation by the fire service at the portals, tunnel service
building(s) or control centre.
The components of FFFS will need to be regularly inspected and tested. There are generally
significant constraints on when such activities can take place, for example closures may be
possible only at night to minimise disruption to traffic. The specification should therefore state
any particular requirements and constraints on where components are to be installed, whether
inside the traffic tubes, cross passages, services gallery or equipment rooms and when equipment
may be maintained.
In existing tunnels, there may also be constraints on the positioning of components of the FFFS
to avoid adverse effects on the existing installations. For example, piping and valve boxes
positioned directly in front of jet fans could impact the efficiency of the ventilation system. Such
conflicts are more likely if the contractor lacks familiarity with the general principles and
practices of road tunnel design and operations.
32
For existing tunnels, site surveys of the tunnel structures, such as tunnel walls, linings and steel
beams, must be done in advance to assess their suitability for the installation of FFFS before the
mechanical fixings for the FFFS are designed, to avoid subsequent problems. In some old tunnels
where asbestos may be present, asbestos surveys must be performed before installation of the FFFS,
with removal or encapsulation of the asbestos required if it may be disrupted by FFFS operation.
5.1.6. Winter Conditions

In countries where severe winter conditions are experienced, the system definition will need to
include provisions to ensure that system operation is not impacted. The provisions could include
the specification of specific materials, insulation of components and trace heating of pipes.
5.1.7. Reliability/Safety Integrity Level

High system reliability is essential. Redundancy principles will generally be specified for critical
elements of the system. Duty and standby units may be specified for the main pump sets, for
example. RAMs analysis may be used to substantiate the likely availability of system components,
taking into account the function of FFFS (whether they are used for asset protection or life
safety), as well as Time Between Failures and Time To Replace data. This should be used to
identify the level of reliability required for the parts of the FFFS. It should be noted that FFFS
are dormant for most of the time and only activated in an emergency or during testing. Therefore,
the term Probability of Failure on Demand (PFD) is therefore more appropriate than Availability
in a RAMs analysis.
The control system for FFFS and other safety critical systems should be subject to strict integrity
requirements. For example, a Safety Integrity Level (SIL) could be specified in accordance with
EN61508. The SIL requirements for safety integrity are based on a probabilistic analysis of the
system. To achieve a given SIL, the system must meet established targets for the maximum PFD.
For example, a SIL 2 system must achieve a PFD of 0.01 or less.
5.1.8. Life Expectancy

The desired life expectancy of FFFS may be stated in the specification or its supporting standards.
A period of the order of 20 to 30 years may be specified as the overall life expectancy of a system
before it is completely replaced, for example. However, this is simplistic because the life
expectancy will vary for the different parts of the system and will depend heavily upon the
maintenance regime. The valves and pumps will deteriorate sooner than the nozzles and the
piping. The control system and its software will probably have a life expectancy only of the order
of 10 years. Some components will require replacement during the lifetime of the system, and
the availability of spare parts should therefore be considered in the specification.
In practice, the life expectancy of the piping and nozzles will be heavily dependent on the
materials used both in the manufacture of the components and their installation in the tunnel. For
components located within the tunnel itself, the requirements for corrosion protection will need
to be stated, notably including provisions to prevent corrosive effects from electrolytic action of
dissimilar metals in contact with each other. The road tunnel standards in numerous countries,
and certain types of high pressure Water Mist Systems, require stainless steel pipes and
components, whereas other countries such as Australia generally use carbon steel pipes.
Thermoplastic coated pipes have been specified for some new tunnels in Sweden.
33
5.2. TESTING AND COMMISSIONING
The requirements for Factory Acceptance Tests (FATs) and Site Acceptance Tests (SATs) for
FFFS will need to be specified. The requirements for tunnel FFFS are similar to those for
suppression systems in general. The requirements for the control system will potentially be more
complex and will need to take into account the interfaces with other systems at the tunnel.
Overall system commissioning will commence upon completion of installation and staged
testing of individual FFFS components and sub-systems. All elements will need to be
commissioned under service conditions.
5.3. APPROVALS
It is important that the requirements and approval process are clearly defined and understood by
all stakeholders at the outset. It would be worthwhile for the stakeholders to meet at the start of
the contract and go through the processes in detail to achieve a good understanding and to
prevent problems at a later stage. The contractor may be unfamiliar with the highway standards
and approval processes applicable for the country and tunnel in question.
Acceptance criteria defined in relation to a performance specification will constitute contractual

obligations. Care is therefore needed to make sure as far as possible that the implications are
understood by the contractor. For example, if the specification states a maximum fire HRR or
maximum temperatures at certain times, then even small exceedances may constitute
unacceptable performance. This can have serious contractual and financial implications for the
parties involved.
5.4. TRAINING
FFFS form an important component of the tunnel safety measures. Training of tunnel staff
involved in operations and maintenance will be essential and will need to be addressed when
specifying and procuring FFFS.
5.5. COST
5.5.1. Capital Cost

The capital cost of a tunnel FFFS can vary widely between tunnels and countries depending on
the circumstances (appendix 1, page44). As outlined above, there are a range of important issues
that need to be considered by the client, contractor and other stakeholders to reduce the
uncertainties that affect cost.
5.5.2. Testing Costs

Full scale fire tests may be required as part of the procurement process, however, the cost of full
scale fire testing is high, which can create commercial pressures that affect the execution, and
potentially the outcome of the contract. Where full scale testing is undertaken, it is important
that adequate allowance is included in the test programme and budget for the live review of
results, and for repeat testing. There is generally only a single opportunity to carry out testing.
It may be virtually impossible to go back to the test tunnel facility at a later stage to repeat tests
if problems are subsequently identified in the results.
34
5.5.3. User Costs

User costs include costs for maintenance, spare parts, routine testing, training of personnel,
operating costs, and energy costs. It is important that full details of the tunnel maintenance
regime and key requirements and constraints are provided at the tender stage so that the FFFS
can be suitably designed to minimise maintenance, and to allow realistic budgetary provisions
to be made.
Maintenance costs for FFFS are dependent on the scope of work that is included. Maintenance
costs typically comprise the tunnel operators management costs, labour costs for regular testing
of the system, and the cost of replacement parts including labour. A number of tunnel operators
around the world were asked to provide details of their maintenance costs. The maintenance cost
for FFFS associated with the labour costs for regular system testing, and replacement parts
including labour (i.e. not including the tunnel operators management costs as noted above), were
approximately 4,200 USD per lane kilometre of the tunnel for the year 2013. Note that the
tunnels surveyed included both older and newer tunnels so that the value provided is representative
for all tunnels. Hidden costs may also result from a tunnel which is closed for maintenance, or
testing and training. These costs may be significant in the case of toll roads and performance
contracts based on high availability of infrastructure.
Other user costs include the training and re-training of persons who might be involved in
inspecting, testing, maintaining or operating FFFS. Therefore, operational costs need to include
the cost of training operations personnel, and if necessary, external agencies such as the fire
safety services and other emergency services. This may include the costs associated with full
scale functional tests such as water, additives, energy, cleaning the tunnel and personnel costs.
Additional information is available in appendix 1, page44.
6. RESEARCH AND ANALYSIS
6.1. RESEARCH PROGRAMS
The year 1999 was a turning point for the consideration of installing FFFS in road tunnels.
The Mont Blanc Tunnel fire and the Tauern Tunnel fire established the potential of tunnel
fires to be larger than previously considered. In 2003, the Runehamar fire tests confirmed
that heavy goods vehicles have a significant potential for large HRRs, up to 200 MW. While
there were other fire tests with FFFS previously, the full scale Water Mist System fire tests
performed for the M30 and A86 projects, in Madrid and Paris respectively, are considered
turning points for considering application of FFFS to road tunnels in Europe and North
America, although FFFS have been installed in Japanese and Australian tunnels for over 30
years.
Fire tests have been traditionally performed to aid researchers in the understanding of the physics
of tunnel fires, understanding the impacts of fires, and verifying calculations, assumptions,
computer models, and tunnel design. Testing and research of FFFS in road tunnels has been
motivated by the need to understand the performance of these systems in the context of fire
development. The primary effort has been to establish the capability of FFFS to protect tunnel
occupants and to allow first responders access to the fire incident. Moreover, the potential for
structural protection has been of interest.
35
Fire tests are primarily performed to verify the ability of FFFS to achieve the protection goals.
Further, fire tests are of vital importance to Tunnel Operators and emergency responders to
coordinate their efforts and verify in practice the emergency response plans.
Some of the key issues are the effectiveness of FFFS, and include various water application rates
and water droplet diameters. Previously, most FFFS testing was performed for buildings and
industrial applications which typically use a smaller fire HRR in design. The issue of how well
FFFS perform under a larger fire HRR has also been assessed via full scale testing.
In the past 10 to 15 years, the design of many new tunnels has considered the installation of
FFFS, even though they are not required. In Appendix 1 there is a survey of tunnels that have
installed FFFS. In the event FFFS are considered, designers and owners have to decide the type
of FFFS to install. This survey shows the variety in systems that are available and installed in
road tunnels. The different types of systems are described in detail in appendix 2, page58.
The test programs that have been performed to verify the performance of FFFS are located in
appendix 4, page71. The broad findings of these programs are also discussed. In summary, the
significant findings in aggregate of these programs are:
FFFS prevent the spread of fire from one target to another;

when there is a stratified smoke layer, FFFS locally force the smoke layer to the roadway;
FFFS reduce visibility within the zones where they are activated, even without fire;
radiation effects from a fire are reduced;
maximum gas temperatures are reduced and the region of tunnel impacted by high heat effects
is significantly minimized;
fire HRR can be reduced;
steam generation is not sufficient enough to be considered a threat.
The technical benefits of FFFS have been demonstrated via full scale testing. FFFS clearly
provide an effective means of managing a fire incident remotely. While FFFS have not been
shown to always extinguish a fire, the evidence shows that FFFS effectively contain a fire
incident and grant the local fire brigade the time and opportunity to approach a fire incident to
address it directly. In addition, the region around a fire incident that may be influenced by high
temperatures is significantly reduced to the fire location.
6.2. MODELLING
Computational Fluid Dynamics (CFD) has been used to model many aspects of tunnel fire safety
for more than 20 years. It can be described as a mature tool to gain better insight into smoke
movement and assess tenability conditions during a fire.
Analysing a problem using CFD requires the modelling of the interaction of solid fuel, liquid
phase and gaseous phase fluids, pyrolysis, combustion, radiation, turbulence and heat transfer.
The area that has the most uncertainty in the CFD modelling of water/fire interaction relates to
the combustion models.
Performing full scale fire tests to assess the optimum parameters for FFFS and their effectiveness
in a fire event is considered the ideal method to verify FFFS performance. Several comparisons
36
and/or validations of various aspects of FFFS operation have been conducted through research
programs and verified using CFD. However, while CFD may be used as a supplement to physical
testing, it is currently not considered to be able to obviate physical testing due to the limitations
of modelling the complex processes occurring during FFFS operation. Refer appendix 5, page75
for a more detailed discussion.
7. CONCLUSIONS AND RECOMMENDATIONS
7.1. CONCLUSIONS
Fire events in tunnels continue to show the significant consequences of these types of events in
a road tunnel environment to tunnel users, the tunnel infrastructure, as well as the impact to the
wider road network on society. This has produced sustained pressure for further improvements
to techniques and technologies to manage the risk and consequence of fires in tunnels. FFFS are
a method that can deliver user safety and infrastructure protection; however, their use is not
widespread for various economic, technical, political and social reasons. This report provides
guidance on the decisions required before adopting FFFS and, if FFFS are to be adopted, provides
guidance on the required design and implementation considerations.
Extensive testing has demonstrated that while FFFS have the ability to reduce the fire HRR and
prevent the fire load reaching its full potential, high gas temperatures may still be reached that
affect the structure or other infrastructure in the immediate vicinity of the fire. This has a direct
link to choosing the correct design fire HRR for the design of FFFS to limit fire growth to, and
the adoption of procedures to assure early activation of systems in the event of fire.
Where installed, maintained and operated effectively, FFFS have a positive impact on egress by
extending the available evacuation time. This benefit applies to vehicles upstream in a
longitudinally ventilated tunnel, and to both sides of a fire in a transversely ventilated tunnel.
However, whilst the conditions downstream of a fire in a longitudinally ventilated tunnel are
significantly improved, untenable conditions may still exist after activation of the FFFS.
The length of tunnel roadway covered by FFFS is affected by the available water supply and the
tunnel width. Operation of FFFS can reduce the visibility for drivers within the area of operation,
however, most vehicles within the activated zone(s) should be stopped as a consequence of the
fire event. Nevertheless, procedures should be adopted to manage traffic and operate the tunnel
systems without exposing motorists to additional hazards. This also means that FFFS should be
reliable and the potential for false activation eliminated.
7.2. RECOMMENDATIONS
7.2.1. Design and Management

Where FFFS are installed, it is essential that they are correctly designed, installed, and integrated
into the tunnel system, as well as properly tested, commissioned, maintained, and operated.
7.2.2. Speed of Operation

FFFS can be activated in the very early stages of fire development before fire fighting activities
commence by trained fire fighters. This allows early suppression and minimises the potential
adverse effects of the fire. Where installed, it is recommended that activation should occur as
37
soon as possible following the fire being detected to minimise fire growth and assure the desired
effectiveness.
7.2.3. Effective Procedures

FFFS should only be activated after confirming the fire location and with the incident vehicle
stopped. Clear plans and procedures are necessary for tunnel operators to activate the FFFS, or
effective automatically operated systems implemented.
7.2.4. Testing and Real Life Incidents

This report provides information on full scale fire testing and the effectiveness of FFFS on real
fire incidents, however, feedback from real incidents has been limited. With the increased use of
FFFS in tunnels, it is important that data of where and how FFFS are operated in the future is
captured and analysed to further develop the understanding and the effectiveness of these systems.
7.3. FUTURE WORK
It is suggested that the following future work or studies could be undertaken to improve the
understanding and performance of FFFS:
human behaviour and responses in a fire environment;

alternative suppression and activation techniques;
emphasis has been on the determination of ceiling temperatures in fire tests and modelling
studies. Further test data and modelling should be undertaken to determine temperatures at
different heights of the tunnel wall during a fire event;
the development of standard test procedures and protocols specific to road tunnels;
the development of a standard methodology for capturing data on fire events and the use of
FFFS so that improvements can be integrated into the system design;
the development of protocols for the specification, commissioning, and proving FFFS;
the development of on-board suppression systems for vehicles.
38
8. BIBLIOGRAPHY / REFERENCES
[1] ARVIDSON, M., Fixed Fire Suppression System Concepts for Highway Tunnels,
International Conference on Tunnel Fires and Escape From Tunnels, pp 129-136, Lyon,
France, 5-7 May, 1999.
[2] ARVIDSON, M., Large-Scale Water Spray and Water Mist Fire Suppression System Tests,
Fourth International Symposium on Tunnel Safety and Security, Frankfurt AM Main,
Germany, pp 283-296 17-19 March, 2010.
[3] BRANDT, A., WIGHUS, R., Real-Scale Tests of Compressed Air Foam System in
Runehamar Test Tunnel 2005, 2006.
[4] CARVEL, R.O., BEARD, A.N., AND JOWITT, P.W., The Influence of Longitudinal
Ventilation Systems on Fires in Tunnels, Tunnelling and Underground Space Technology,
16, pp 3-21, 2001.
[5] CARVEL, R.O., BEARD, A.N., JOWITT, P.W., AND DRYSDALE, D.D., Variation of
Heat Release Rate with Forced Longitudinal Ventilation for Vehicle Fires in Tunnels, Fire
Safety Journal, 36, 6, pp 569-596, 2001.
[6] CARVEL, R.O., BEARD, A.N., JOWITT, P.W., The Influence of Longitudinal Ventilation
and Tunnel Size on HGV Fires in Tunnels, 10th International Fire Science and Engineering
Conference (Interflam 2004), pp 815-820, Scotland, 5-7 July, 2004.
[7] GRITZO L, et al, The Influence of Risk Factors on Sustainable Development, FM Global,
chapter 3.7, 2009.
[8] INGASON, H., LONNERMARK, A., Effects of Longitudinal Ventilation on Fire Growth
and Maximum Heat Release Rate, Fourth International Symposium on Tunnel Safety and
Security, Germany, pp 395-406, 17-19 March 2010.
[9] INGASON, H., Model Scale Tunnel Fire Tests Longitudinal Ventilation, SP Swedish
National Testing and Research Institute, SP Report 2005:49, Boras, Sweden, 2005
[10] JNSSON, J., JOHNSON, P., Suppression System - Trade-Offs and Benefits, Fourth
International Symposium on Tunnel Safety and Security, pp 271-282, Frankfurt AM Main,
2010.
[11] KNAPP, D., http://dknappfiredesign.com.au/interesting_tunnels_page.html.
[12] LEMAIRE, A., D., MEEUSSEN, V., Effects of Water Mist on Real Large Tunnel
Fires: Experimental Determination of BLEVE - Risk and Tenability During Growth and
Suppression, Efectis Nederland Report, R0425, June 2008.
[13] LEMAIRE, T., MEEUSSEM, V., Experimental Determination of BLEVE - Risk Near
Very Large Fires in a Tunnel with a Sprinkler / Water Mist System, Fourth International
Symposium on Tunnel Safety and Security, Frankfurt AM Main, Germany, pp 283-296,
17-19 March, 2010
[14] LNNERMARK, A., On the Characteristics of Fires in Tunnels, Doctoral Thesis,
Department of Fire Safety Engineering, Lund University, Lund Sweden, 2005.
[15] LNNERMARK, A., KRISTENSSON, P., HELLTEGEN, M., BOBERT, M., Fire
Suppression and Structure Protection for Cargo Train Tunnels: Macadam and HotFoam,
3rd International Symposium on Safety and Security in Tunnels (ISTSS 2008), pp 217-228,
Stockholm, Sweden, 12-14 March, 2008.
[16] LNNERMARK, A., INGASON, H., The Effect of Air Velocity on Heat Release Rate
and Fire Development During Fires in Tunnels, 9th International Symposium on Fire Safety
Science, pp 701-712, Karlsruhe, Germany, 21-26 September 2008
[17] MAWHINNEY, J. R., TRELLES, J., Computational Fluid Dynamics Modelling of Water
Mist Systems on Large HGV Fires in Tunnels, presented at the Journe dEtude Technique:
39
Brouillard dEau Quoi de Neuf, at Ple Europen de Scurit CNPP - Vernon, France, 22
November, 2007.
[18] MAWHINNEY, J., Fixed Fire Protection Systems in Tunnels: Issues and Directions, Fire
Technology, 2011.
[19] NATIONAL FIRE PROTECTION ASSOCIATION, NFPA 13 Standard for the Installation
of Sprinkler Systems, National Fire Protection Association, 1 Batterymarch Park, PO Box
9101, Quincy, MA 02269-9101, USA, 2010 Edition.
[20] NATIONAL FIRE PROTECTION ASSOCIATION, NFPA 502 Standard For Road
Tunnels, Bridges, And Other Limited Access Highways, National Fire Protection Association,
1 Batterymarch Park, PO Box 9101, Quincy, MA 02269-9101, USA, 2014 Edition.
[21] NATIONAL FIRE PROTECTION ASSOCIATION, NFPA 750 Standard on Water Mist
Fire Protection Systems, National Fire Protection Association, 1 Batterymarch Park, PO
Box 9101, Quincy, MA 02269-9101, USA, 2010 Edition.
[22] OTA, O., Automatic Fire Extinction (Sprinkler) System, OTA Engineering, Tokyo, Japan,
December 2002.
[23] SOLIT, Engineering Guidance for a Comprehensive Evaluation of Tunnels with Fixed Fire
Fighting Systems Scientific report of the SOLIT research project, SOLIT2 Consortium, 2012.
[24] SOLIT2, Safety of Life in Tunnels 2, Engineering Guidance for a Comprehensive Evaluation
of Tunnels with Fixed Fire Fighting Systems, Scientific Final Report of The Solit2 Research
Project, Prepared by the Solit Research Consortium, 2012
[25] STROEKS, R., Sprinklers in Japanese Road Tunnels, Bouwdienst Rijkswaterstaat,
Directoraat-Generaal Rijkswaterstaat, Ministry of Transport, Netherlands, Project Report
BFA-10012, 2001.
[26] UPTUN Engineering Guidance for Water Based Fire Fighting Systems for the Protection of
Tunnels and Subsurface Facilities, European Commission research project UPTUN, Work
Package 2, Fire Development and Mitigation Measures, Report D251, September 2008.
[27] UPTUN, Workpackage 2 Fire development and mitigation measures - D251: Engineering
Guidance for Water Based Fire Fighting Systems for the Protection of Tunnels and Sub
Surface Facilities, 2008.
[28] UPTUN, Workpackage 2 Fire development and mitigation measures - D253: Summary of
Water Based Fire Safety Systems in Road Tunnels and Sub Surface Facilities, 2008.
[29] WORLD ROAD ASSOCIATION (PIARC), PIARC Technical Committee C5 on Road
Tunnels, Fire and Smoke Control in Road Tunnels, 05.05.B - 1999.
[30] WORLD ROAD ASSOCIATION (PIARC), Technical Committee C3.3 Road Tunnel
Operations, Road Tunnels: an Assessment of Fixed Fire Fighting Systems, 2008R07.
Operations, Systems and Equipment for Fire and Smoke Control in Road Tunnels, 05.16.B
- 2007.
Operations, Road Tunnels: Operational Strategies for Emergency Ventilation, 2011R02.
[33] CETU, Water Mist in Road Tunnels State of knowledge and provisional assessment
elements regarding their use, Information Document, 2010
[34] APSAD, Rfrentiel APSAD R1: Extinction automatique eau de type sprinkler (in French),
2014
[35] BAST 2014a: Wirksamkeit automatischer Brandbekmpfungsanlagen in Straentunneln,
(in German), Project No. FE 15.0563/2012/ERB, Bundesministerium fr Verkehr und
digitale Infrastruktur (BMVI), Bundesanstalt fr Straenwesen (BASt), written by ILF
Consulting Engineers in cooperation with Universitt Regensburg and STUVA e.V., 2014
40
[36] BAST 2014b: Wirtschaftlichkeit automatischer Brandbekmpfungsanlagen in Straentunneln,

(in German), Project No. FE 15.0564/2012/ERB; Bundesministerium fr Verkehr und digitale
Infrastruktur (BMVI), Bundesanstalt fr Straenwesen (BASt), written by ILF Consulting
Engineers in cooperation with Amstein + Walthert AG and STUVA e.V., 2014
[37] WORLD ROAD ASSOCIATION (PIARC), PIARC Technical Committee C4 Road
Tunnel Operations, Risk Evaluation, Current Practice for Risk Evaluation for Road Tunnels,
2012R23EN.
[38] Directive 2004/54/EC of the European Parliament and of the Council of 29 April 2004 on
minimum safety requirements for tunnels in the Trans-European Road Network.
[39] WANG, A.X., RHODES, N., KOTTAM, K., TRAPANI, R., CFD Simulations of Ventilation
Effects on Water Mist Fixed Fire Suppression Systems on Tunnel Fires, 15thInternational
Symposium on Aerodynamics, Ventilation and Fire in Tunnels, pp 677-686, Barcelona,
Spain, 18-20 September, 2013.
[40] DIX, A., http://www.scribd.com/doc/122942826/Dix-Report-Burnley-Tunnel-Incident-2007.
[41] BECHTEL/PARSONS BRINCKERHOFF, Memorial Tunnel Fire Test Ventilation
Program, Comprehensive Test Report, Prepared For Massachusetts Highway Department/
Federal Highway Administration, 1995.
[42] CESMAT, E., et al., Assessment of Fixed Fire-Fighting Systems for Road Tunnels by
Experiments at Intermediate Scale, 3rd International Symposium on Tunnel Safety and
Security, Stockholm, Sweden, 2008
[43] CHEONG, M., CHEONG, W., LEONG, K., LEMAIRE, A., TARADA, F., Heat
Release Rates of Heavy Goods Vehicle Fires in Tunnels, 15th International Symposium on
Aerodynamics, Ventilation and Fire in Tunnels, Barcelona, Spain, 18-20 September 2013.
[44] DEL REY I., ESPINOSA I., FERNANDEZ S., GRANDE A., ALARCON E., Ventilation
System Design and Large Scale Fire Tests, 4th International Conference on Tunnel Safety
and Ventilation, Graz, Austria, 2008
[45] Directorate-General for Public Works and Water Management, Civil Engineering Division,
Project Safety Test Report On Fire Tests, Directorate-General for Public Works and
Water Management-Civil Engineering Division, Griffioenlaan 2, P.O. Box 20000, 3502 La
Utrecht, August 2002
[46] FERNANDEZ, S., DEL REY, I., GRANDE, A., ESPINOSA, I., ALARCON, E., Large
Scale Fire Tests for the Calle 30 Project, 5th International Symposium on Tunnel Safety
and Security, 2012, New York.
[47] FERNANDEZ, S., DEL REY, I., GRANDE, A., ESPINOSA, I., ALARCON, E.,
Ventilation and FFFS Fire Tests for Calle 30 Road Tunnels, 15th International Symposium
on Aerodynamics, Ventilation and Fire in Tunnels, Barcelona, Spain, 18-20 September,
2013.
[48] GUIGAS, X., et al., Dynamic Fire Spreading and Water Mist Test for the A86 East Tunnel,
5th International Conference on Tunnel Fires, London, UK, 25-27 October, 2004.
[49] HAERTER, A., Fire Tests in the Ofenegg-Tunnel in 1965, International Conference on
Fires in Tunnels, SP Report 1994:54, p 195-214, Bors, Sweden, 10-11 October, 1994.
[50] HEJNY, H., Task 2.3: Evaluation of Current Mitigation Technologies in Existing Tunnels,
Technical Report, UPTUN, 2006.
[51] INGASON, H., Fire Testing in Road and Railway Tunnels, in Flammability Testing of
Materials used in Construction, Transport and Mining, V. Apted, Editor, Woodhead
Publishing pp 231-274, 2006
[52] INGASON H, Model Scale Tunnel Tests With Water Spray, Fire Safety Journal 43 (7):pp
512-528, 2008.
41
[53] KAWABATA, N., KUNIKANE, Y., YAMAMOTO, N., TAKEKUNI, K., SHIMODA, A.,
Numerical Simulation of Smoke Descent in a Tunnel Fire Accident, 4th International
Conference on Tunnel Fires, p 357-366, Basel, Switzerland, 2002.
[54] KO, YOON, A Study of the Heat Release Rate of Tunnel Fires and the Interaction between
Suppression and Longitudinal Air Flows in Tunnels, Doctoral Thesis Presented to the Department
Of Civil and Environmental Engineering at Carleton University, Canada, April 2011
[55] Kommision fr Sicherheitsmassnahmen in Strassentunneln, chlussbericht der Versuche im
Ofenegg Tunnel Von 17.5 31.5 1965, 1965.
[56] KUNIKANE, Y., KAWABATA, N., OKUBO, K., SHIMODA, A., Behaviour of Fire
Plume in a Large Cross Sectional Tunnel, 11th International Symposium on Aerodynamics
and Ventilation of Vehicle Tunnels, p 78 - 93, Luzern, Switzerland, 2003.
[57] KUNIKANE, Y., KAWABATA, N., ISHIKAWA, T., TAKEKUNI, K., SHIMODA, A.,
Thermal Fumes and Smoke Induced by Bus Fire Accident in Large Cross Sectional Tunnel,
The 5th JSME-KSME Fluids Engineering Conference, Nagoya, Japan, 17-21 November,
2002.
[58] KUNIKANE, Y., KAWABATA, N., TAKEKUNI, K., SHIMODA, A., Heat Release
Rate Induced by Gasoline Pool Fire in a Large-Cross-Section Tunnel, 4th International
Conference on Tunnel Fires, 387-396, Basel, Switzerland, 2-4 December, 2002.
[59] OPSTAD, K., STENSAAS, J.P. BRANDT, A. W., Task 2.4: Development of New
Innovative Technologies, Technical Report, UPTUN, 2006.
[60] SHIMODA, A., Evaluation of Evacuation Environment During Fires in Large-Scale
Tunnels, 5th Joint Workshop COB/JTA p 117 - 125, Japan, 2002.
[61] SOLIT, KRATZMEIR, S., Safety of Life in Tunnels, Forschungsbericht Solit -
Wassernebelanlagen in Strassentunneln (Only available in German), 2007
[62] TUOMISAARI, Maarit, Full-Scale Fire Testing For Road Tunnel Applications Evaluation
Of Acceptable Fire Protection Performance, Proceedings from the Third International
Symposium on Tunnel Safety and Security, Stockholm, Sweden, SP Technical Research
Institute Of Sweden, 12-14 March 2008
[63] INGASON, H., APPEL, G., LI, YZ, Large scale fire test using Fixed Fire Fighting System
in Runehamar Tunnel, SP Report 2014:32
[64] BLANCHARD, E., BOULET, p., CARLOTTI, P., Capability of a CFD Tool for Assessing a
Water Mist System in a Tunnel, 15th International Symposium on Aerodynamics, Ventilation
and Fire in Tunnels, pp 717-727, Barcelona, Spain, 18-20 September, 2013
[65] VAARI, j., HOSTIKKA, S., SIKANEN, t., PAAJANEN, A., Numerical Simulations on the
Performance of Water-Based Fire Suppression Systems, VTT Technology 54, 2012
[66] Comite europeen de Normalisation Publications, EN 12845 Fixed Firefighting Systems -
Automatic Sprinkler Systems - Design, Installation and Maintenance, CEN-CENELEC
Management Centre, Avenue Marnix 17 - B-1000 Brussels, 2004 Edition
[67] UPTUN Engineering Guidance for Water Based Fire Fighting Systems for the Protection of
Tunnels and Subsurface Facilities, European Commission research project UPTUN, Work
Package 2.5, Report R251, August 2006.
[68] WORLD ROAD ASSOCIATION (PIARC), Technical Committee C4 on Road Tunnel
Operations, Human factors and road tunnel safety regarding users, 2008R17.
[69] KOELL, M., Single Tunnel and still safe The Felbertauern Tunnel, 4th International
Conference on Tunnel Safety and Ventilation, pp 133-138, Graz, Austria, 21-23 April, 2008
[70] LEUCKER, R., LEISMANN, F.M., Fire Tests for Water Mist Fire Suppression Systems
in Road Tunnels, Underground. The Way To The Future, Proceedings Of The World Tunnel
Congress, pp 366375, Geneva, Switzerland, 31 May - 7 June, 2013.
42
[71] LEMAIRE., A.D., MEEUSSEN, V.J.A., Experiments to determine BLEVE-risk and

downwind tenability conditions during growth and suppression of real large tunnel fires,
Efectis Nederland BV: A TNO Company, March 2008.
[72] TARADA, F., NOORDIJK, L.M., CHEONG, M.K., CHEONG, W.O., LEONG, K.W., The
Energy Budget in Suppressed Tunnel Fires, 15th International Symposium on Aerodynamics,
Ventilation and Fire in Tunnels, Barcelona, Spain, 18-20 September 2013.
[73] CHAN, E., TARADA, F., Crossing Points, Fire Risk Management Journal, Fire Protection
Association, UK, March 2009.
[74] INGASON, H., LI, Y.Z., LNNERMARK, A., Handbook of Tunnel Fire Dynamics,
chapter 16, pp 403-441, Fire Suppression and Detection in Tunnels, Springer, 2015.
[75] INGASON, H., LI, Y.Z., Technical Trade-Offs Using Fixed Fire Fighting Systems,
pp 90-97, 7th International Conference on Tunnel Safety and Ventilation, Graz, Austria,
12-13 May 2014.
[76] INGASON, H., APPEL, G., LI, Y.Z., Large Scale Fire Tests with Fixed Fire Fighting
System in Runehamar Tunnel, SP Report, 2014:32, SP Technical Research Institute of
Sweden.
43
9. GLOSSARY
TERM DEFINITION
Where smoke from a fire flows in the direction opposite to the ventilation system airflow. (See also
Backlayering
Critical Velocity).
Boiling Liquid The explosive release of expanding vapour and boiling liquid following the catastrophic failure, due to
Expanding Vapour an external fire, of a pressure vessel holding a pressure liquefied gas such as LPG (liquefied petroleum
Explosion gas) or CNG (compressed natural gas). Note: the event may result in fireballs, blast, projectiles and
(BLEVE) possible toxic vapour cloud. (Abbr. BLEVE).
Compressed Air A pressurised water system that allows the addition of compressed air to produce a stream of foam
Foam System during pumping operations. (Abbr. CAFS).
Computational
The use of numerical methods and algorithms to solve and analyse 3-dimensional fluid flow problems.
Fluid Dynamics
(Abbr. CFD).
(CFD)
Concrete Traffic Concrete impact barriers to protect tunnel infrastructure. Note that a New Jersey Barrier (NJB) is a
Barrier type of Concrete Traffic Barrier. (Abbr. CTB).
The velocity of air within a longitudinally ventilated tunnel that creates sufficient airflow to oppose
Critical Velocity the buoyancy forces of hot smoke from a fire and thereby prevents the smoke from backlayering in the
direction opposite to the ventilation system airflow. (Abbr. Vc).
An open, FFFS comprising relatively large water droplets activated on a zone-by-zone basis. Note:
Deluge System
the operation of the system can be automatic or manual.
The primary pipeline providing pressurised fire fighting water (or other fluids) along a tunnel for final
Fire Main
distribution to section valves, hydrant valves, and other suppression systems via standpipes.
Heat Release Rate The rate of heat energy output from a fire. (Abbr. HRR).
Heavy Goods Heavy Goods Vehicles are trucks and buses where the truck can be a single truck, lorry-trailer
Vehicle combinations, articulated trucks, or semi-trailers. (Abbr. HGV).
Hydrant A coupling connected to a standpipe so that fire hoses can be connected.
A factor used to calculate the volumetric flow rate of water discharge for each nozzle as a function of
K-factor
nozzle operating pressure.
Probability of
A value, expressed as a percentage, that indicates the probability of a system failing to respond on
Failure on
demand. It equates to 100% minus the percentage Safety Availability. (Abbr. PFD).
Demand
Reliability, Availability and Maintainability (RAM) is a modelling methodology used to simulate and
Reliability,
predict a systems capabilities for a given configuration in terms of its operation, failure modes and
Availability, and
failure rate, maintainability and hence its overall availability. RAMs analysis is used as a comparative
Maintainability
tool to compare the performance of different system configurations. (Abbr. RAM).
A valve or device that can be activated to open or close the water supply from a fire main to the
Section Valve pipework providing water to a zone downstream. Activation is generally via a control system either
automatically or manually. Some jurisdictions may also require activation to be able to occur locally.
A connecting pipe or adaptor that enables water to be delivered from a supply point or main to a point
of connection. For fire fighting purposes, the point of connection is generally a hydrant point,
Standpipe
sprinkler outlet or section valve. In a tunnel environment the standpipe is dedicated to fire purposes.
Standpipes can be either wet or dry.
An open, FFFS comprising relatively fine water droplets activated on a zone-by-zone basis. Note: the
Water Mist System
operation of the system can be automatic or manual.
Zone A section of tunnel served by a section valve for a deluge of water mist system.
44
10. APPENDICES
APPENDIX 1. QUESTIONNAIRE DATA
A1.1 Introduction
In 2012, a questionnaire was sent to numerous road tunnel agencies around the world to gather
data on the use of FFFS. Responses were received and collated from mid 2012 to early 2014.
52 responses were received and they are believed to provide a representative sample. The output
of the questionnaire provides some guidance to those who may seek to use FFFS. The results are
presented below.
A1.2 Tunnels without FFFS

Although there were only eight respondents in this category, the vast majority of road tunnels
worldwide have no FFFS installed. As reinforced by the responses, the predominant reasons for
not using FFFS are due to the age of the tunnel, in that FFFS were not required at the time of
tunnel construction, and there was no advantage seen in the use of FFFS as insufficient data
were available in the tunnel community on the benefits and operational criteria to justify its use.
A1.3 Tunnels considering the use of FFFS in the future

Reasons for FFFS being considered in the future for the five respondents appears to predominantly
be that application of FFFS provides a capital advantage in that some other systems do not need
to be provided in the tunnel system with the application of FFFS, and that it is seen that FFFS
provide enhanced life safety and protection to the tunnel structure. Insufficient data were
available to determine if FFFS of one particular type were preferred. The type of system is
generally to be determined by future studies at the particular tunnels noted.
A1.4 Tunnels using water mist

The nine tunnels that responded to the questionnaire use high pressure Water Mist System with
a water application rate of between 0.5 to 0.73 l/min/m3. Design fire sizes for this application rate
vary from 30 MW to 200 MW. The predominant reasons for using FFFS were for life safety, and
protection of the tunnel structure and equipment. It is interesting to note that in most of the
tunnels listed, the potential to apply cost offsets for using FFFS was not a relevant determinant
for using FFFS.
A1.5 Tunnels using deluge

The 24 tunnels that responded to the questionnaire use Deluge Systems with a water application
rate of between 6 to 12 mm/min. Design fire sizes for this application rate vary from 20 MW to
200 MW. Tunnels with a higher design fire size generally had a higher water application rate
implying that some consideration was given to the design fire size when selecting the application
rate. The predominant reasons for using FFFS were for life safety, and protection of the tunnel
structure and equipment. The use of FFFS to offset the cost of implementation of other systems
appears to be almost equally split where some choice was available to the tunnel designer (it is
noted that the implementation of a Deluge System in Australia is mandatory).
A1.6 Tunnels with other systems installed

Responses were received from two tunnels that do not have a Water Mist System or Deluge
System (i.e. the system installed is not a system that complies with the limitations listed in
chapter 1.2). It is further noted that a number of tunnels that have a Water Mist System or a
45
Deluge System installed also have other types of fixed water supply systems including standpipes
and hydrants.
Appendices A1.7 to A1.11 page 46 to 57, are presented as tables.

46
2016R03EN
A1.7 TUNNELS WITHOUT FFFS

DENMARK /
COUNTRY DENMARK UK UK UK UK UK UK
SWEDEN
National State Highway A102 Blackwall A102 Blackwall
TUNNEL Oresund A101 Rotherhithe A13 Limehouse Link A38 Saltash Clyde
Network Northbound Southbound
All existing road East London East London East London
Between Denmark and
LOCATION tunnels in the Danish (Thames River (Thames River (Thames River East London Cornwall Glasgow
Sweden
State Network Crossing) Crossing) Crossing)
Bi directional (cars
TRAFFIC FLOW Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Tidal Flow Unidirectional
and vans only)
Varies, mainly 2 lane
LANES PER TUBE 2 2 2 2 3 Single bore, 3 2
but some 3 lane
LENGTH (m) 200 to 1,000 4,000 1,483 1,350 1,174 1,553 410 760
TUNNEL SHAPE Rectangular Rectangular Horseshoe Horseshoe Horseshoe Rectangular Horseshoe
VENTILATION
Longitudinal Longitudinal Semi-transverse Longitudinal Longitudinal Longitudinal Semi transverse
SYSTEM
FFFS not required at FFFS not required at
a) Tunnel not
time of construction: time of construction:
required to have
a) Not considered a) Not considered
FFFS not required FFFS not required FFFS not required FFFS at time of
cost-effective; cost-effective;
at time of at time of at time of FFFS not required at construction;
b) The system may not b) The system may not
construction. construction. construction. time of construction. b) Lack of clarity of
extinguish the fire; extinguish the fire;
REASONS FOR NO Insufficient current Insufficient Insufficient current Insufficient current benefits of FFFS;
c) Parts of the tunnel c) Parts of the tunnel
FFFS data on benefits current data on data on benefits data on benefits and c) Semi-transverse
are protected with a fire are protected with a
and operational benefits and and operational operational criteria nature of the
insulating material; fire insulating material;
criteria to justify operational criteria criteria to justify to justify use. ventilation system;
d) The local fire d) The local fire
use. to justify use. use. d) Maintenance
brigade can be at the brigade can be at the
costs.
fire scene within 10 fire scene within 10
minutes. minutes.
FIXED FIRE FIGHTING SYSTEMS IN ROAD TUNNELS
A1.8 TUNNELS CONSIDERING THE USE OF FFFS IN THE FUTURE
COUNTRY AUSTRALIA AUSTRIA SINGAPORE SINGAPORE UK UK UK
Sentosa Gateway North South
TUNNEL Legacy Way S16 Arlberg Conwy Conwy Pen-y-clip Pemaenbach
Tunnel Expressway Tunnel
LOCATION Brisbane, QLD Singapore Singapore Wales A55 Wales A55 Wales A55
Unidirectional (bi- Unidirectional Unidirectional (bi-
Unidirectional, Unidirectional,
TRAFFIC FLOW Unidirectional Bi-directional directional during (bi-directional during directional during
one tube two separate tubes
maintenance) maintenance) maintenance)
LANES PER TUBE 2 2 2 3/4 2 2 2
LENGTH (m) 13,972 840 12,300 1,090 930 658
TUNNEL SHAPE Rectangular Rectangular Rectangular Rectangular Hybrid Hybrid
VENTILATION
Longitudinal Full Transverse Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal
SYSTEM
a) Life safety;
a) Life safety; a) Life safety; a) Life safety;
b) Protection of the a) Life safety; a) Life safety;
REASONS FOR b) Protection of the b) Protection of the b) Protection of the
tunnel structure and b) Protection of the b) Protection of the
PLANNING TO Mandated tunnel structure; tunnel structure; tunnel structure;
escape routes; tunnel structure. tunnel structure.
INSTALL FFFS c) Cost offset for other c) Cost offset for other c) Cost offset for other
c) Cost offset for other
safety systems. safety systems. safety systems.
safety systems.
a) Passive fire a) Passive fire a) Passive fire a) Passive fire

protection; protection; protection; protection;
b) Optimisation of b) Increase to b) Increase to b) Increase to
ventilation system; To mitigate the need for To mitigate the need for ventilation system; ventilation system; ventilation system;
COST OFFSET Not applicable
c) Improved escape passive fire protection passive fire protection c) Fire rating of M&E c) Fire rating of M&E c) Fire rating of M&E
egress; equipment; equipment; equipment;
d) Fire brigade d) Improved escape d) Improved escape d) Improved escape
response time. egress. egress. egress.
TYPE OF FFFS
Water mist with AFFF
PLANNED TO BE Deluge System Deluge System Deluge System To be determined To be determined To be determined
extinguishing media
USED
Tanks with 1 hour Tanks with 1 hour
storage with continuous storage with continuous Town main with tank Town main with tank Town main with tank
WATER SOURCE Tanks Tanks
direct feed from water direct feed from water back-up back-up back-up
main main
Automatic Incident Automatic Incident
Linear Heat Detection Detection, CCTV, Detection, CCTV,
DETECTION
and Video Incident Linear Heat Detector; Linear Heat Detector; To be determined To be determined To be determined
METHOD
Detection manually activated on manually activated on
confirmation of a fire confirmation of a fire
47
2016R03EN
48
2016R03EN
A1.8 TUNNELS CONSIDERING THE USE OF FFFS IN THE FUTURE

COUNTRY AUSTRALIA AUSTRIA SINGAPORE SINGAPORE UK UK UK
45 seconds for time of
ACTIVATION TIMES Within 4 minutes Within 4 minutes
detection to discharge
150 MW (HGVs
allowed in tunnel but
30 MW (buses, vans not trailers and
and cars are only dangerous goods
MAX ANTICIPATED allowed to enter the vehicles, HGV fire size
FIRE SIZE tunnel, bus fire size based on free burning
adopted from NFPA large scale fire test in
502 2011 Table A11.5.1) Spain conducted by
Singapore Land
Transport Authority)
100 MW for ventilation
50 MW for ventilation; 30 MW for ventilation
DESIGN FIRE SIZE 50 MW design if deluge system
200 MW for structure design
installed
CAPITAL COST EU 14 M
MAINTENANCE
EU 700 k
COST / YEAR
A1.9 TUNNELS USING WATER MIST
COUNTRY AUSTRIA AUSTRIA FINLAND NETHERLANDS NETHERLANDS SPAIN UK UK UK
Helsinki City
TUNNEL Felbertauern Mona Lisa Roertunnel Swalmen Calle 30 Dartford Tyne - NB Tyne - SB
Service Tunnel
River Thames,
Dartford
LOCATION Matrei Linz Helsinki Roermond Swalmen Madrid Wallsend Wallsend
Thurrock
Crossing
TRAFFIC FLOW Bi-directional Bi-directional Bi-directional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional
LANES PER TUBE 2 2 2 2 2 3 to 4 2 2 2
Protected length
LENGTH (m) 5,300 800 2,200 2,015 429 1,435 1,650 1,500
= 2,352
TUNNEL SHAPE Rectangular Horseshoe Horseshoe Rectangular Rectangular Rectangular Hybrid Hybrid Hybrid
VENTILATION Longitudinal and
Fully transverse Longitudinal Longitudinal Longitudinal Longitudinal Semi-transverse Longitudinal Longitudinal
SYSTEM semi-transverse
a) Life safety;
b) Protection of
a) Protection of As a consequence As a consequence
tunnel structure
the tunnel of administrative of administrative a) Life safety;
and tunnel a) Life safety; a) Life safety;
a) Protection of structure; agreements and to agreements and to a) Protection of b) Protection of

equipment; b) Protection of b) Protection of
the tunnel b) Protection of set up a pilot set up a pilot the tunnel the tunnel
c) Protection of the tunnel the tunnel
REASONS FOR structure; assets built over program to gain program to gain structure; structure and
external traffic structure; structure;
INSTALLING FFFS b) Traffic of the tunnel experience with experience with b) Protection of tunnel
network; c) Protection of c) Protection of
Dangerous (railway line); FFFS. FFFS. tunnel equipment;
d) Protection of tunnel tunnel
Goods in tunnel. c) Traffic of Provide protection Provide protection equipment. c) Cost offset for
assets built over equipment. equipment.
Dangerous for the tunnel for the tunnel other systems.
the tunnel;
Goods in tunnel. structure. structure.
e) Cost offset for
other systems.
a) Passive fire
a) Optimise
protection;
ventilation
b) Optimisation
system;
of ventilation
b) Savings in
COST OFFSET Not applicable Not applicable Not applicable Not applicable Not applicable system; Not applicable Not applicable
drainage and
c) Improved
water supply as
escape egress
well as water
d) Fire brigade
treatment.
response time.
High pressure High pressure High pressure High pressure High pressure High pressure High pressure High pressure High pressure
TYPE OF FFFS water mist water mist water mist water mist water mist water mist water mist water mist water mist
(30 - 40 bar) (30 - 40 bar) (80 bar) (30 - 40 bar) (30 - 40 bar) (50 - 140 bar) (45 bar at nozzle) (50 bar at nozzle) (50 bar at nozzle)
49
2016R03EN
50
2016R03EN
A1.9 TUNNELS USING WATER MIST

COUNTRY AUSTRIA AUSTRIA FINLAND NETHERLANDS NETHERLANDS SPAIN UK UK UK
WATER SOURCE Tanks Town Main Town Main Tanks Tanks Tanks Tanks Tank Tank
SPRAY DENSITY
0.78 0.70 0.50 0.78 0.80 0.70 0.72 0.70 0.73 0.50 0.50
(l/min/m3)
25 m long, width 25m long, width
Full tunnel width Full tunnel width Full tunnel width Full tunnel width x Full tunnel width x variable variable
ZONE DIMENSIONS 14 m x 24 m 7 m x 25 m
x 36 m long x 40 m long x 25 - 30m long 25 m long 25 m long depending on depending on
tunnel width tunnel width
2.8 m to 3.6 m 2.8 m to 3.6 m
SPACING OF
2m 2m 3.5 m 2m 2m 3.2 m 3.57 m dependant on dependant on
DISCHARGE HEADS
tunnel section tunnel section
ZONES OPERATED
3 3 2 3 3 3 3 3 3
SIMULTANEOUSLY
DURATION OF FFFS
5 hours 2 hours 1 hour 1 hour 1 hour 1 hour 1 hour
DISCHARGE
Controlled by Controlled by
Linear Heat Linear Heat Linear Heat Linear Heat SCADA. Input SCADA. Input
Detection and Linear Heat Linear Heat Linear Heat Detection and Detection and Detection and from VAID, Heat from VAID, Heat
DETECTION METHOD
Video Incident Detection Detection Detection Video Incident Video Incident Video Incident Detector, Manual Detector, Manual
Detection Detection Detection Detection Call Point, Call Point,
Tunnel Controller Tunnel Controller
Activation by Activation by Activation by Discharge is 120 seconds or 120 seconds or
ACTIVATION TIMES Manual
operator on Manual operator on operator on initiated by an Manual earlier if initiated earlier if initiated
(detection to activation activation by fire
detection and activation detection and detection and indication from activation by Tunnel by Tunnel
time) brigade
verification verification verification fire-fighters Controller Controller
ACTIVATION TIMES
Approx. 60 Approx. 60 Approx. 60
(time from system 10 seconds 10 seconds 20 30 seconds 10 seconds 10 seconds 30 seconds
seconds seconds seconds
activation to discharge)
One delivery
MAX ANTICIPATED
200 MW 200 MW truck (max 12 m 200 MW 200 MW 60 MW 100 MW 200 MW 200 MW
FIRE SIZE
long)
DESIGN FIRE SIZE 200 MW 200 MW 20 MW 200 MW 200 MW 30 MW 200 MW 200 MW
Design based on Fire test not Fire test not
full scale fire Functionality test Fire test to undertaken. undertaken.
Acceptance Acceptance
tests undertaken Full functionality Full functionality undertaken as 45MW Results based on Results based on
COMMISSIONING criteria = VdS / criteria = VdS /
as part of the test test well as a fire test undertaken in a results from results from
IBS Certificate IBS Certificate
system to 90 MW test tunnel SOLIT research SOLIT research
development project. project.
Approx. 1.5 M Approx. 4.1 M

CAPITAL COST Approx. 6 M EU Approx. 15.5 M EU Approx. 6.3 M EU Approx. 5 M EU 8.1 M GBP
EU EU
MAINTENANCE COST With TyneSB, With TyneNB,
30 k EU 20 k EU 249 k EU 158 k EU 75 k EU
/ YEAR 17 k GBP 17 k GBP
A1.10 Tunnels using deluge systems
Due to the number of responses from tunnels with Deluge Systems, the data are presented in three tables two for Australian Tunnels (tables 5 and 6) and one
for tunnels in other countries around the world (table 7, page 55).
TABLE 5 - TUNNELS WITH DELUGE SYSTEM INSTALLED AUSTRALIA (1 OF 2)

Sydney Harbour Cross City Eastern Inner City
TUNNEL M2 Lane Cove Tunnel M5 East Tugan Bypass Clem 7
Tunnel Tunnel Distributor ByPass
Arncliffe,
LOCATION Sydney, NSW Epping, NSW Lane Cove, NSW Sydney, NSW Tugan, NSW Sydney, NSW Brisbane, QLD Brisbane, QLD
NSW
TRAFFIC FLOW Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional
LANES PER TUBE 2 2 2 2 2 2 2 2
LENGTH (m) 2,300 800 3,600 4,000 2,100 400 1,700 4,800 350
TUNNEL SHAPE Rectangular Horseshoe Horseshoe Horseshoe Rectangular Rectangular Rectangular Horseshoe Rectangular
VENTILATION
Semi transverse Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal
SYSTEM
REASONS FOR
Mandated
INSTALLING FFFS
COST OFFSET Not applicable
Town Mains and
WATER SOURCE Town Mains Town Mains Town Mains Town Mains Town Mains Town Mains Town Mains Town Mains
tanks
SPRAY DENSITY
12 10 10 10 10 10 10 10 7.5
(mm/min)
30m long x 25 m long x 25 m long x 25 m long x tunnel 30m long x
ZONE DIMENSIONS 8 m x 45 m 10.5 m x 50 m 32,346 m 2
tunnel width tunnel width tunnel width width tunnel width
SPACING OF
3m 5m 5m 5m 3m 3m 3m 50 m
DISCHARGE HEADS
ZONES OPERATED
2 2 2 2 2 2 2
SIMULTANEOUSLY
DURATION OF FFFS
Continuous Continuous Continuous 1 hour Continuous Continuous Continuous 1 hour 1 hour
DISCHARGE
Automatic Automatic
Incident Incident Automatic Incident Linear Heat
Linear heat Detection, Detection, Detection, CCTV, Detection, Video
Linear Heat
detector plus CCTV, Linear CCTV, Linear Linear Heat Incident
Detector or Manual activation
DETECTION METHOD Flame detector video Heat detector; Heat detector; detector; manually Detection, Frangible bulbs
manual by operator
monitoring by manually manually activated on Visibility and
activation
the operator activated on activated on confirmation of a CO measuring
51
confirmation of confirmation of fire system

a fire a fire
2016R03EN
52
2016R03EN

Sydney Harbour Cross City Eastern Inner City
TUNNEL M2 Lane Cove Tunnel M5 East Tugan Bypass Clem 7
Tunnel Tunnel Distributor ByPass
ACTIVATION TIMES
(detection to activation
time)
ACTIVATION TIMES
(time from system
MAX ANTICIPATED
FIRE SIZE
DESIGN FIRE SIZE 50 MW
Hydrants
installed as well
OTHER NOTES
as a Deluge
System
COMMISSIONING
CAPITAL COST
MAINTENANCE
COST / YEAR
George Mullum Northbridge Boggo Road Northern
TUNNEL Domain Tunnel Burnley Tunnel Airport Link Heysen Tunnel
Bridges Tunnel Mullum Tunnel Tunnel Busway Busway
LOCATION Nundah, QLD Melbourne, VIC Melbourne, VIC Melbourne, VIC Perth, WA Brisbane, QLD Adelaide, SA Brisbane, QLD Brisbane, QLD
TRAFFIC FLOW Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Unidirectional Bi-directional Bi-directional
LANES PER TUBE 2 2
LENGTH (m) 285 1,600 1,800 3,200 500 5,300 500 500
Rectangular and Rectangular and Rectangular and
TUNNEL SHAPE Horseshoe Rectangular Rectangular Horseshoe Rectangular
Horseshoe Horseshoe Horseshoe
VENTILATION
Longitudinal Longitudinal Hybrid Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal
SYSTEM
REASONS FOR
Mandated
INSTALLING FFFS
COST OFFSET Not Applicable
WATER SOURCE Town Mains Town Mains Town Mains Town Mains Town Mains Town Mains Tanks Town Mains
SPRAY DENSITY
7.5 7.5 7.5 7.5 7.5 10 7.5 6 6
(mm/min)
ZONE DIMENSIONS
SPACING OF
DISCHARGE HEADS
ZONES OPERATED
SIMULTANEOUSLY
DURATION OF FFFS
1 hour 1 hour 1 hour 1 hour 1 hour 1 hour 1 hour 1 hour 1 hour
DISCHARGE
Automatic
Incident
Detection,
CCTV, Linear
DETECTION METHOD Heat detector;
manually
activated on
confirmation of
a fire
ACTIVATION TIMES
(detection to activation
time)
53
2016R03EN
54
2016R03EN

George Mullum Northbridge Boggo Road Northern
TUNNEL Domain Tunnel Burnley Tunnel Airport Link Heysen Tunnel
Bridges Tunnel Mullum Tunnel Tunnel Busway Busway
ACTIVATION TIMES
(time from system
MAX ANTICIPATED
FIRE SIZE
DESIGN FIRE SIZE 50 MW 30 MW
OTHER NOTES
COMMISSIONING
CAPITAL COST
MAINTENANCE COST
/ YEAR
TABLE 7 - TUNNELS WITH DELUGE SYSTEM INSTALLED OTHER COUNTRIES
COUNTRY JAPAN KOREA KOREA MEXICO SINGAPORE SINGAPORE SPAIN
Woodsville Marina Coastal De Vielha Juan
TUNNEL Happusan JuckRyung InJae Acapulco
Interchange Tunnel Expressway Carlos I
DanYang-goon, InJae-goon,
LOCATION Acapulco Singapore Singapore Valle de Aran (Lerida)
Kyoungbuk-Do Gangwon-do
TRAFFIC FLOW Unidirectional Unidirectional Unidirectional Bi-directional Unidirectional Unidirectional Bi-directional
LANES PER TUBE 2 2 2 3 2 5 3
LENGTH (m) 4,471 4,600 10,965 2,947 260 3,600 5,230
TUNNEL SHAPE Horseshoe Horseshoe Horseshoe Horseshoe Rectangular Rectangular Rectangular
VENTILATION SYSTEM Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Longitudinal Hybrid
a) Life safety;
b) Protection of the
a) Protection of the a) Protection of the
tunnel structure;
tunnel structure; tunnel structure;
c) Protection of tunnel
b) Protection of tunnel a) Life safety; b) Protection of tunnel a) Life safety;
REASONS FOR equipment;
equipment; Life safety b) Protection of the equipment; b) Protection of the Life safety
INSTALLING FFFS c) Protection of
c) Protection of tunnel structure. c) Protection of tunnel structure.
external traffic
external traffic external traffic
network;
network. network.
d) Protection of assets
built in the tunnel.
Optimisation of
ventilation system,
To mitigate the need
Fire Brigade time of passive fire protection,
COST OFFSET Not applicable Not applicable Not applicable for passive fire Not applicable
response spacing of cross
protection
passages, Fire Brigade
response time.
TYPE OF FFFS Deluge System Deluge System Deluge System Deluge System Deluge System Deluge System Deluge System
Tanks with 45 hour Tanks with 1 hour
storage with storage with
WATER SOURCE Tanks Tank Tank Tank Tank
continuous direct feed continuous direct feed
from water main from water main
21 (currently
SPRAY DENSITY reviewing to reduce to
6 6 6 12 6
(mm/min) 12 following large
scale fire tests)
ZONE DIMENSIONS 9.2 m x 50 m 8 m x 45 m 10.5 m x 50 m 32,346 m 2 25 m x tunnel width 25 m x tunnel width 100 m x tunnel width
SPACING OF
5m 5m 5m 150 m 3m 3m 50 m
DISCHARGE HEADS
55
2016R03EN
56
2016R03EN
TABLE 7 - TUNNELS WITH DELUGE SYSTEM INSTALLED OTHER COUNTRIES

COUNTRY JAPAN KOREA KOREA MEXICO SINGAPORE SINGAPORE SPAIN
ZONES OPERATED
2 2 2 23 2 2 6
SIMULTANEOUSLY
DURATION OF FFFS 10 hours with 2
40 minutes 20 minutes 40 minutes Continuous Continuous 20 minutes
DISCHARGE nozzles
Automatic Incident Automatic Incident Linear Heat Detection,
Linear heat detector Detection, CCTV, Detection, CCTV, Video Incident
Linear Heat Detector Manual activation by
DETECTION METHOD Flame detector plus video monitoring Linear Heat detector; Linear Heat detector; Detection, Visibility
or manual activation operator
by the operator manually activated on manually activated on and CO measuring
confirmation of a fire confirmation of a fire system
ACTIVATION TIMES
(detection to activation 3 minutes (automatic) 10 to 15 minutes Within 4 minutes Within 4 minutes
time)
ACTIVATION TIMES
(time from system A couple of seconds 5 seconds
200 MW (HGVs 200 MW (HGVs
allowed but not trailers allowed but not trailers
MAX ANTICIPATED 2 vehicle car or truck and dangerous goods and dangerous goods
Large sized bus 100 MW 50 MW 30 MW
FIRE SIZE incident vehicle. HGV fire size vehicle. HGV fire size
adopted from NFPA adopted from NFPA
502 Table A11.5.1) 502 Table A11.5.1)
0.9 m 2 fire pan with 6 vehicle car or truck 200 MW for 200 MW for
DESIGN FIRE SIZE 20 MW 50 MW 30 MW
gasoline incident ventilation design ventilation design
Hydrant System
OTHER NOTES installed as well as
Deluge System
Fire test not Fire test not Fire test not
Fire test using 0.5 m 2 Fire test not Fire test not undertaken. undertaken. undertaken.
COMMISSIONING 5 MW pool fire
fire pan with gasoline. undertaken undertaken Functionality test Functionality test Functionality test
undertaken. undertaken. undertaken
CAPITAL COST 800 M YEN 200 M EU
MAINTENANCE COST /
8 M YEN 20 M USD 4.6 M EU
YEAR

A1.11 TUNNELS WITH OTHER SYSTEMS INSTALLED
COUNTRY CHINA CHINA
TUNNEL Baiyun Rd Xianghuang Rd
LOCATION Chongqing Xianghuang Rd, Nanan District, Chongqin
TRAFFIC FLOW Bi-directional Unidirectional
LANES PER TUBE 2 2
LENGTH (m) 7,100 890
TUNNEL SHAPE Horseshoe Horseshoe
VENTILATION SYSTEM Longitudinal / Hybrid Longitudinal
a) Life safety;
a) Life safety;
b) Control fire growth;
REASONS FOR INSTALLING FFFS b) Protection of tunnel structure;
c) Extinguish fire;
c) Protection of assets built over the tunnel.
d) Traffic of dangerous goods in tunnel.
COST OFFSET Not applicable Not applicable
TYPE OF FFFS Foam Hydrant system
WATER SOURCE Tanks Tank
SPRAY DENSITY 20 l/s 5 l/s
DURATION OF FFFS DISCHARGE 6 hours 180 minutes

DETECTION METHOD Heat Detection, Video incident Detection, manual inspection Manual activation
ACTIVATION TIMES
3 minutes (automatic)
(Detection to discharge time)
ACTIVATION TIMES
A couple of seconds
Time from system activation to discharge
MAX ANTICIPATED FIRE SIZE 20 MW
DESIGN FIRE SIZE 20 MW 20 MW
OTHER NOTES
COMMISSIONING Fire test not undertaken, functionality test undertaken. Fire test not undertaken
CAPITAL COST 800 M YEN
MAINTENANCE COST / YEAR 8 M YEN
57
2016R03EN
58
APPENDIX 2 - TYPES OF SYSTEMS
A2.1 Introduction
Although the term FFFS includes all types of fire-fighting systems in tunnels there is a huge
variation in the names and capabilities of different systems. Traditional Deluge Systems are
those found in Japan and Australia, consisting of conventional deluge nozzles in zones that can
be activated either automatically, or by the operator from the tunnel operation control room.
Water Mist Systems may also be designed with zones, the same is true for foam systems, although
the zone length may differ. As defined in chapters 1.4.1 and 1.4.2, page4, Deluge System will
be used for deluge systems without foam additives, Water Mist System for systems operating at
both high and low pressure, and Foam for all foam systems using the addition of air, including
High Expansion (Hi-Ex) and Compressed Air Foam Systems (CAF).
A2.2 Deluge Systems

A2.2.1 General Description
Deluge Systems consist of open sprinklers or deluge nozzles attached to pipework at the tunnel
ceiling. The pipework consists of mains pipes, manifold pipes, feed mains and branch pipes. The
sprinklers or nozzles are attached to the branch pipes, which are typically arranged in a uniform
pattern at the ceiling to distribute spray to all sections of the roadway. The branch pipes are
connected to a feed main which is connected to a section valve. The section valve is mounted on
a manifold attached to the mains pipe that is supplied by one or more water reservoirs or fire
pump stations. Mains pipes are normally filled with water up to the point of connection to the
section valve (wet); therefore the mains pipe and the section valves must be protected against
freezing where this climatic condition exists. The section valve separates the wet mains pipe
from the empty (dry) feed main and branch pipes supplying the sprinklers or spray nozzles.
When the section valve is opened, water flows into the feed main and branch pipes and discharges
from the open sprinklers.
Illustration 7 Section valve arrangement in cross passage

Clem 7, Brisbane, Australia
The branch piping is divided into deluge zones, typically 25 m to 50 m in length, each served by
its own section valve. An independent fire detection system that is capable of locating a fire
59
accurately is required, so that the section valve serving the zone where the fire is located can be
released. The section valve can be opened either automatically by the detection system, or
manually by a signal from the tunnel operator. If an incident occurs on the boundary between
two deluge zones, both zones will need to be activated. When the section valve opens, water
flows into the feed main and branch piping and discharges from all sprinklers or nozzles in that
deluge zone. As the deluge nozzle (or sprinkler head) orifices are open, the branch piping is at
atmospheric pressure until water is introduced. A deluge system has a time delay between
detection of a fire and the discharge of water from the sprinklers or nozzles due to the time
required to operate the valve (which will depend on whether activation is automatic or manual)
and to fill the branch piping network with water and reach the desired operating pressure.
The system should be designed with sufficient water capacity to allow simultaneous operation of
at least two consecutive deluge zones, but depending on the precision provided by the detection
system, it may be necessary to design for three operating zones, one in the incident area, and the
adjacent upstream and downstream zones. The length of the deluge zones should be coordinated
with the pumping capability as well as the fire detection and ventilation zones. Piping should be
designed to allow drainage of water from all piping between the section valve and the sprinklers
or nozzles after the flow has been stopped.
Deluge Systems are specified based on the area of coverage required. Due to the constraints and
configuration of the section valves, this is generally in the order of 300 m2, however, this value
varies according to various national and international standards. The width laterally across the
tunnel needs to consider where vehicles will end up after a collision. Where Concrete Traffic
Barriers (CTBs) are used, the lateral dimension is considered to be from the toe of the CTB to
the toe of the opposite CTB. Where low kerbs are present, the width needs to be extended to the
tunnel lining. This concurs with recommendations provided in NFPA 502 [20], and current
standard installation practice. NFPA502[20] states that deluge nozzles should be spaced such
that the system coverage extends to the roadway shoulders and if applicable, maintenance and
patrol walkways.
The other considerations in the design of deluge systems are the length of the longest vehicle to
ensure the entire vehicle is drenched, and the possibility that a vehicle may stop at the interface
between two deluge zones. This means that the minimum number of deluge zones that are
required to be activated simultaneously for any tunnel is two.
Deluge discharge densities in existing facilities vary between 2.5 mm/min to 12 mm/min. 6 mm/
min is used as the basic requirement for Japanese tunnels, 6 mm/min has been used on bus only
tunnels in Australia and 7.5 mm/min to 12 mm/min has been used for road tunnels in Australia.
The design discharge densities are a minimum and therefore need to be calculated so that they
are achieved in the most hydraulically disadvantaged location.
The current minimum duration for deluge discharge is 40 minutes in Japan and 1 hour in
Australia. Note that to maintain the required deluge density, any water volume requirement for
hydrant operation will need to be added to the deluge requirements.
A2.2.2 Specific Technical Information

The length of deluge zones typically varies from 25 m to 50 m. Standard deluge nozzles, which
typically require a minimum operating pressure of 1.5-5 bar, are used and they discharge a
60
uniform pattern of water droplets over the protected area with droplet sizes less than 1-2 mm in
diameter. The K-factor of the nozzles is typically 80 l/min/bar1/2. Tests with fires having potential
free burning heat release rates in the order of 25 MW to 150 MW have been undertaken with
deluge systems [74].
The most suitable length of the deluge zones must be based on the width of the tunnel and the
capacity of the water supply. Large zones will reduce the number of control valves but place a
higher total water demand [2]. This may mean that at gore areas of the tunnel (i.e. merges and
diverges) where the tunnel width increases, the length of the deluge area may need to be decreased
to keep to the standard nominal spray area and to keep the water supply requirement manageable.
The typical application rates and zone sizes can result in flow demands in the range of 7,500 to
15,000 litres per minute, which can have a significant impact on supply and drainage system
requirements [3]. This value is very much dependent on the tunnel width. For example in a 15 m
wide tunnel, a density of 10 mm/min, and a two operating deluge zones of 50 m would require
15,000 litres/min. If the tunnel width is 10 m instead, the corresponding flow would be
10,000 litres/min, which is significantly lower.
If an incident occurs on the boundary between two zones, both zones will need to be activated.
The type of fire alarm initiating device is selected mainly based on the hazard (e.g. smoke
detectors, heat detectors, CCTV or optical flame detectors). The initiation device signals the fire
alarm panel, which in turn signals the section valve to open. Activation can also be manual,
depending on the fire protection objectives of the system. Manual activation is usually done via
an electric or pneumatic fire alarm pull station, which signals the fire alarm panel, which in turn
signals the section valve to open. According to the SOLIT guidelines [23], deluge systems may
be activated and operated manually or automatically depending on the availability of trained
personnel, the risks expected, the type of deluge system, the control systems used and applicable
legislation. NFPA 502 [20] recommends that the time delay should not exceed three minutes in
order to prevent the development of a major fire.
The UPTUN [26] and SOLIT [23] guidelines recommend that the installation of pumps comply
with the manufacturers documented requirements. Pumps shall be installed in a dedicated pump
room or other designated area. Adequate ventilation and drainage shall be provided. The pump
room shall be lockable to prevent access of unauthorized personnel. Deluge System shall be
designed to provide at least 110% of the nominal flow required for the most demanding protection
area in the tunnel. This shall be calculated at the minimum nozzle pressure as type tested in full
scale fire tests. The required flow rate shall be provided by one or more pumps.
According to the SOLIT guidelines [23], the duration time shall be determined in a specific risk
analysis for every individual tunnel. The system shall be capable of a minimum activation time
of 30 minutes, although longer activation times are normally required. A minimum of 60 minutes
shall be used for tunnels longer than 500 m, however, in practice 90 to 120 minutes are probably
necessary to account for the response capabilities of the fire department.
A2.2.3 Examples of Deluge Systems

Australia
Australia has installed Deluge Systems into its road tunnels since the Sydney Harbour Tunnel was
built. This tunnel was opened in 1992. Currently there are 19 tunnels with deluge systems in operation.
61
In Australia, a section valve station is generally located every 120 m along the tunnel length.
This location coincides with the location of cross passages or egress passages and therefore the
valves are located inside a fire rated space. The section valves are designed so that operation is
automatic during a fire event but also allows the operator to be able to open and close section
valves as required during a fire scenario, especially if the fire moves or spreads.
The deluge zone length can vary but has generally been designed around a deluge zone area of
300 m2 which covers the full width of the roadway. Consequently, the length of the deluge zone
can vary according to tunnel width. The system is designed for simultaneous activation of
however many zones are required to provide complete coverage of the maximum length vehicle
that uses the tunnel plus allowance to cover the possibility that the vehicle may be at the boundary
of two zones. Current common practice is to provide a water discharge density of between 7.5 to
10 mm/min in road tunnels. Australia also has some tunnels that are only used by buses. The
water discharge density for these tunnels is generally 6 mm/min. In a fire scenario, the system
flow rate is designed to operate for 60 minutes at full flow while a number of hydrants operate
simultaneously. Pumps and tanks (if required) are duplicated so that no single failure can affect
the Deluge System performance.
Activation of the deluge system is usually by manual operation from a remote control room. The
operator receives an alarm from one or a number of detection systems such as a Video Automatic
Incident Detection (VAID) system, linear heat detection system, other Closed Circuit TV (CCTV)
cameras and/or manual alarm calls. On receipt of the alarm, the operator confirms that there is
a fire event and activates the Deluge System. Most systems are configured so that on alarm,
unless the operator intervenes, the Deluge System activates. However, the operator can initiate
the system prior to automatic operation. The operational intent is to activate the Deluge System
as soon as possible while the fire is still small (i.e. less than 10-20 MW).
Japan
Japan introduced Deluge Systems into its high risk expressway tunnels 45 years ago and currently
there are over 120 systems in operation. Different technical solutions are applied, depending on
the owner of the tunnel. The Japanese Deluge Systems are designed for 6 mm/min. The pressure
at the nozzle location is between 3 and 3.5 bars. There are either 50 m spray zones or 100 m spray
zones. Depending on the owner there are different distances between nozzles in each zone.
Water reservoir capacity should be designed as 40 minutes for the operation time for two deluge
zones (50 m or 100m) [25]. System design and operation is as follows:
fire detectors are located on the tunnel side wall at 25 to 50 m spacing within the whole section
for initial detection of the fire;
the fire location is confirmed in the control room by CCTV, at which point the Deluge System
is manually activated for a 50 m zone around the seat of the fire until the fire brigade arrives
at the fire site; and
to minimize the risk of fire spread, one additional deluge zone will be activated.
Technically the Japanese water and foam sprinkler systems are automatic in design in combination
with fire detector and automatic valve control. However, as automatic operation of sprinkler
could cause a traffic accident, the tunnel operator must recognize the fire and confirm its
existence by CCTV, before starting the sprinkler system. Once the fire has been visually
confirmed, the sprinkler system is started manually as quickly as possible [22].
62
USA
While FFFS are not commonly provided in tunnels in the USA, the City of Seattle Fire Department
has made it a requirement to install FFFS in road tunnels since the installation of FFFS in the
Battery Street Tunnel (State Route 99) in 1954. The Deluge System installed in the Battery Street
Tunnel is operated by spot heat detectors which operate weight actuated mechanical valves and
are further supported by the provision for manual and remote operation. This is one of the oldest
FFFS installations in the world and, even though it is 60 years old, still functions using nearly all
of its original infrastructure.
The Seattle Fire Department specifies the use of a positive alarm sequence process for FFFS
activation. Positive alarm sequence provides both automatic activation in the correct zone(s),
and allows the tunnel operator the choice of delaying activation while the alarm source is
investigated by other sources such as CCTV.
To date, all FFFS installed in the USA are deluge, and in addition to the Battery Street Tunnel,
FFFS are installed in:
Interstate 5 (I5) Seattle, 0.8 km in 12 lanes (foam deluge) (1988);

Mt Baker Tunnel Interstate 90 (I90) - Seattle, 3.9 km in 3 bores (foam deluge) (1993);
Downtown Seattle Transit Tunnel Seattle, bus tunnel (deluge) (1992);
Downtown Seattle Transit Tunnel Seattle, retrofit for the addition of passenger trains (deluge)
(2009); and
Port of Miami Tunnel - Miami (deluge) (2014).
FFFS are also currently being installed, or planned to be installed, in the following projects:
Presidio Parkway Tunnels San Francisco (deluge) estimated completion date 2016;
Virginia Midtown Tunnel Norfolk (deluge) estimated completion date 2016;
SR99 Alaska Way Viaduct Replacement Seattle, 1.6 km stacked tunnel (2 lanes in each
direction) (foam deluge) - estimated completion date 2016;
I-90 (3 bores) 2.5 miles - Seattle (foam deluge) estimated completion date 2018; and
SR 520 1 mile Seattle, 6 km (unknown FFFS) estimated completion date 2022.
Sweden
The Swedish Traffic Administration installed a simplified Deluge System in the Northern Link
Tunnel. An improved version of this concept is also planned to be used in the Stockholm Bypass
when it will open in 2020. In total, the system will be installed in 50 km of tunnels.
The design is based on considerations regarding simplicity, robustness, investment costs and
maintenance issues. To meet these design requirements the system consists of:
63
a single pipe in the centre line of the tunnel ceiling, fitted with two extended coverage nozzles
(large K-factor nozzles) directed horizontally towards each of the tunnel walls. The entire
cross section of the 14 m wide tunnel is covered with only one pipe. The nozzle used for the
Northern Link Tunnel has a K factor of 240 l/min/bar1/2 and the nozzle for the Stockholm
Bypass will have a K factor of 360 l/min/bar1/2;
long sections of 50 m to 75 m (deluge zones) are used and are designed for delivering 5-10mm/
min without the use of any additives to the water. If two sections have to be activated due to a
fire between two sections a lower water density is accepted;
the Deluge System is combined with the fire hydrant system, reducing the standard requirement
for two water mains in the tunnel;
the water supply is obtained by connection to the public water supply, and no additional pumps
are required. This means that the discharge time is virtually unlimited;
thermoplastic coated steel pipes and clamp couplings.
The main purpose of the system is to limit the fire size and prevent fire spread during the
evacuation period in congested traffic situations. When the traffic is flowing freely the need for
the system is regarded as minor. The system can be manually operated from the traffic control
centre based on detection by CCTV, or from the tunnel escape routes where the section valves
are located. The system also starts automatically if a heat sensing cable detects high temperatures
from a fire. The sprinkler pipes are self-draining due to the risk of freezing. In winter the
temperature in the traffic space is expected to drop below -20 C.
A2.3 Water mist systems

A2.3.1 General Description
Water Mist Systems are fundamentally similar to Deluge Systems (i.e. the pipework consists of
a water filled mains pipe, manifold, section valves, dry feed main and branch pipes to which the
nozzles are attached to provide water to a specific zone). The mains pipe is connected to a water
supply and pumps generate the pressure. Water Mist Systems may vary with respect to their
working pressures, i.e. low pressure and high pressure systems. The piping or tubing utilized in
the system must be designed for the corresponding operating pressure. To protect against
plugging of small orifice nozzles, water mist systems utilize corrosion resistant materials such
as stainless steel pipe or tubing. The primary difference between the systems is the percentage
of smaller droplet sizes, as the droplet size is inversely proportional to the pressure applied, and
the momentum of the spray ejected from the nozzles, as spray from high pressure nozzles
typically has higher momentum than spray from low pressure nozzles.
The effectiveness of the mist system is based on a strong correlation between spray density,
momentum and droplet size. Mist systems are based on volume of coverage. As mist extends to
the total volume, the area calculated needs to consider the tunnel width to the wall lining, the
zone length, and the tunnel height. As an example, water densities of 0.5 to 1.5 l/min/m3 with
different control objectives have been tested and implemented in some relevant projects (A86 in
Paris, M30 in Madrid). As an example, for a zone width of 10 m and zone length of 30 m, tunnel
height of 6 m and assuming a water density rate of 1 l/min/m3, with 2 zones operating, the above
requirements mean that a minimum flow rate of 3,600 l/min is required.
A2.3.2 Specific Technical Information

The definition given in the UPTUN guidelines [26] states that the general principle of the low
pressure water mist system is to produce a fog (or mist) of small water droplets at a nozzle
64
pressure of 3-10 bar. The high pressure water mist system produces a fog (or mist) with a mix of
different sizes of water droplets at a nozzle pressure of 60-120 bar.
From the Appendix table in the UPTUN document [26], the total water flow rate per 25 m zone
for low pressure systems (without additives) is in the range of 221-683 l/min, and for high
pressure systems, 140-550 l/min. Note, however, that the total water flow rate depends on the
tunnel width, zone length and the number of zones operating. One zone of 25 m in a 10 m wide
tunnel, at 2.3mm/min, would require 575 l/min. Designing for two zones would require a
pumping capacity of 1150 l/min (+ 10% of the nominal flow required), and if for three 25 m
zones, 1.725 l/min (+ 10%). If the tunnel is more than 10 m wide and the zones longer than 25 m,
the hydraulic demand and pump capacity is much higher. The discharge rate for low pressure
systems is in the range of 1.1-3.3 mm/min (l/min/m2) and for high pressure systems 0.5-2.3 mm/
min. Note that the design application densities are based on a density per unit area of coverage
(mm/min). They are sometimes converted to other measures when discussing water mist systems,
namely a volumetric density expressed as a flow rate per volume (l/min/m3) by simply dividing
mm/min by the ceiling height of the tunnel in metres. This means that densities are identical
when expressed in terms of tunnel area, but very different when expressed in terms of volume
when comparing two tunnels with the same width, but totally different tunnel heights. The
K-factor for a high pressure nozzle can vary between 4.0 - 5.5 l/min/bar1/2. The length of each
zone can vary from 20 m to 25 m and up to three zones can be activated at once.
Water Mist Systems use significantly less water than Deluge Systems. On the other hand they
require significantly higher pressure, especially the high pressure system. As a result, pipes,
tanks and pump capacities can be smaller, and the water demand be lowered. Likewise, drainage
volumes can potentially be lowered [10].
The SOLIT guidelines [23] state that a high pressure Water Mist System applies nozzle pressures
above 35 bars. Low pressure Water Mist Systems apply nozzle pressure of less than 12 bars. The
medium pressure Water Mist Systems apply nozzle pressure between 12-35 bar. Water Mist
Systems apply small water droplets as the fire-fighting agent. The diameter of droplets from a
water mist nozzle are usually less than 1 mm, leading to a Dv0.90 of 1 mm, meaning that 90%
of the volume of the spray consists of droplet sizes of less than 1 mm. NFPA 750 [21] uses a
Dv0.99 value instead of a Dv0.90 value to define a water mist.
Centrifugal pumps are typically used for low pressure and medium pressure systems, whereas
positive displacement (PD) pumps (or assemblies of PD pumps) are typically used for medium
and high pressure systems. The pump capacity is selected to provide at least 110% (+10%) of the
nominal flow required for the most demanding protection area in the tunnel. This is calculated
at the minimum nozzle pressure as type tested in full scale fire tests. The minimum output
capacity for positive displacement pumps, or assemblies of PD pumps, is 90 l/min. The minimum
capacity for centrifugal pumps is 750 l/min. The water tank is suitable for providing water for all
simultaneously activated sections (typically two or three) with the required flow rate based on
the defined minimum period of operation [23].
There are water mist systems available which are designed to share the riser pipe and pump units
with the hydrant system.
65
A2.3.4 Examples of Water Mist Systems

Tyne Tunnel Newcastle, United Kingdom
The Tyne tunnels in the Newcastle upon Tyne region cross beneath the River Tyne linking the
town of Jarrow on the south bank with North Shields and Howden on the north side of the Tyne.
The crossing consists of a pedestrian and cyclist tunnel opened in 1951 as well as two road
tunnels opened in 1967 as part of the A19. The New Tyne Road Tunnel was opened in 2010. The
volume of traffic using the tunnel currently amounts to 38,000 vehicles per day with a predicted
increase to 43,000 vehicles per day by 2021.
The New Tyne Tunnel project has become a pioneer in the field of tunnel fire fighting in Europe
through the decision to install FFFS to suppress fire for protecting a road tunnel 1.7 km in
length. It more than conforms to current legislation in the United Kingdom as well as European
standards. The decision to install a fire suppression system in the New Tyne Crossing project
was reached following a recommendation by experts based on a quantitative risk assessment and
a cost benefit analysis [73]. According to the study, the mean benefit-cost ratio for the installation
of FFFS was 1.27, primarily through the expected reduction in traffic delays following a fire.
Both the new road tunnel and the original tunnels are fitted with FFFS on a water mist basis. The
protected areas are divided into a total of 130 zones each 25 m long. In the event of fire, three
neighbouring zones are activated at the same time [24].
Dartford Tunnel, United Kingdom

The Dartford River Crossing Tunnels are located some 25 km from the centre of London and
link Dartford on the south bank of the Thames with Thurrock on the north side. The Thames
crossing as part of the M25 London motorway ring comprises two road tunnels and the Queen
Elizabeth II Bridge and is used by around 150,000 vehicles per day. The two tunnels are altogether
1.43 km in length. The first tunnel was opened for traffic in 1963, the second followed in 1981.
The Highways Agency decided to retrofit a stationary high-pressure Water Mist System from
2010 to 2012, to enhance protection for motorists, the service staff, and emergency services in
the event of fire [24].
A86 Paris, France

The A86, also known as the super-priphrique or Priphrique de lIle-de-France, is a motorway ring
around Paris some 78 km in length. The final section completing the ring in the west is designed as a
duplex car tunnel with the driving lanes on top of each other and is equipped with FFFS.
The A86 Tunnels total length amounts to roughly 10,300 m and it was opened to traffic in 2008.
Taking the related connecting tunnels and the two driving levels into account, around 24 km of
tunnel bores had to be provided with fire fighting measures. The A86 Tunnels special feature is
that it possesses a circular cross-section driven by a tunnel boring machine, which is used for
two traffic levels. This results in a clear ceiling height for the traffic levels of only 2.55 m and
consequently no heavy goods vehicles are allowed to use the tunnel. Three driving lanes are
available per traffic level and direction with one-way traffic. The two traffic levels are linked to
one another every 200 m by means of stairwells and there is an evacuation route leading into the
open every 1,000 m. The tunnel is equipped with a ventilation system, which provides fresh air
to the two traffic levels through special air ducts. Air exhaust and fresh air ducts run along the
ceiling and in the area below the carriageway.
66
The suppression system in the A86 Tunnel, with a total of 850 zones and 16,000 spraying heads,
was applied under real conditions for the first time in December 2010, when a car caught fire.
The system worked as planned and the tunnel was able to return to busy as usual only one and
a half hours after the fire broke out. All those involved assessed the findings obtained during this
deployment as extremely positive [24].
Felbertauern Tunnel, Austria

The Felbertauern tunnel, opened in 1967, has a length of 5,300 m and is operated as a single tube
tunnel with bi-directional traffic. As this tunnel goes under the main ridge of the Alps it faces
quite often high pressure differences between the portals. This results in natural air velocities of
8 to 10 m/s without any support by ventilation or moving vehicles. Such high air velocities pose
a high risk should a fire event occur. A fully transverse ventilation system provides sufficient
ventilation in normal operation for air quality and smoke extraction for a fire event. As the
tunnel was not equipped with a means of egress, the refurbishing process in 2005/2006 focused
on improving the safety standard and increasing self-rescue by introducing additional egress
methods. Traffic volume is too low to justify the construction of a second tube. Due to the high
overburden and the long length of the tunnel, the construction of a parallel running egress gallery
was also not considered due to cost.
To improve the safety of the tunnel, the decision was made to use the existing fresh air duct,
located above the traffic envelope, as the egress route. At the time, this was a novel solution and
required major civil works. Staircases were erected approximately every 230 m to connect the
road level with the new egress route.
Illustration 8 Sketch of tunnel with staircase to the Egress route

Koell M., Single Tunnel and Still Safe - The Felbertauerntunnel:
Proceedings of the 4th Symposium on Tunnel Safety and Ventilation, April 2008 Graz
67
One important new safety feature was the installation of a high pressure Water Mist System. This
system was intended to reduce the risk to tunnel users by reducing the HRR and size of a fire and
to protect the intermediate ceiling, above which the new egress route runs. It therefore fulfils the
goals of both life and structure protection. A high pressure Water Mist System was chosen due to
its benefits in water consumption and the reduced space requirements. In the event of an incident
three adjacent zones, each with a length of 36 m, will be activated. The water mist application rate
is approximately 4 l/min/m, which equates to 3,800 l/min in the three activated zones.
Due to challenging environmental conditions in wintertime, the system performance has to be

met for temperatures down to -30C at the portals. Consequently, special frost protection
measures as well as stringent requirements on the materials where required [69].
Vielha Tunnel, Spain

The new Vielha Tunnel, opened in December 2007, was built to replace an existing tunnel,
which will only be used as an evacuation tunnel and for transporting hazardous goods in the
future. The new tunnel is 5.2 km in length, with varying gradients (550 m with +1.7% and
4,550 m with -4.5%). It operates with a two-way, three lane system with a total width of 14 m.
The ventilation system operates as a semi-transverse flow system and is divided into four
ventilation sections. These are connected to ventilation stations at the portals. Cross passages
lead to the original tunnel for evacuation purposes are arranged at 400 m spacings.
The FFFS are solely activated by manual means by the tunnel operations centre via remote
control. At present, activation is only foreseen once all tunnel users and the emergency services
have vacated the fire zone. In addition, activation of the fire fighting system is coordinated with
the fire detection and ventilation system [24].
M30 Tunnel - Madrid, Spain

The M30 forms Madrids inner motorway ring and represents one of Europes largest urban road
tunnel projects undertaken so far and has a length of approximately 56 km. The project was
constructed between September 2004 and summer 2007. Parts of the tunnel as well as technical
operational rooms are protected by water mist systems [24].
A2.4 Foam Systems

There are mainly three types of foam systems. A foam Deluge System with injected foam
concentrates into the water supply, a High Expansion Foam System (Hi-Ex) and Compressed Air
Foam (CAF).
A foam Deluge System is a specific application system, discharging low expansion foam,
resulting in a foam spray from the sprinkler. Foam Deluge Systems are effective at controlling
fires involving flammable liquid spills in tunnels, but they are also effective on conventional
lorry fuel loads [12]. Systems using injected foam concentrates can be both Deluge Systems and
Water Mist Systems as described in chapter 1.2, page3.
The discharge density needed in order to extinguish or control flammable liquid fires using
water with a film forming additive is reasonably well established. Information is for example
given in NFPA 16, which recommends an average discharge density of 6.5 mm/min. Large scale
fire suppression tests in tunnels show good performance of a foam-water sprinkler system. Water
and foam additive, 3% AFFF, was pumped from a container to the deluge zone with nozzles.
68
Tests show that the effectiveness of the foam deluge system was not negatively affected by a
longitudinal ventilation velocity of 4.2 m/s. The test fires were extinguished in less than
30 seconds [13].
Technology involving CAF [14] or Hi-Ex [15] has been tested against both solid and liquid fuel
fires. These foam system tests demonstrated a good degree of fire control. As pointed out by
Mawhinney[17] neither the CAF nor Hi-Ex systems have been widely accepted for use in
tunnels. One reason is uncertainty about the effect of loss of visibility on fire fighting and rescue
operations, particularly with Hi-Ex foam.
APPENDIX 3 - MAINTENANCE AND TESTING
A3.1 Introduction
Maintenance of FFFS includes all technical, operational and administrative requirements so that
the system performs to the specified performance during its design life. Regular maintenance of
the FFFS commences immediately after the initial testing, commissioning and proving of the
system.
The design life of the system can be defined as the period from the start of operational use until
the system needs total replacement. It should be noted that some components of the system may
need to be replaced a number of times during the design life of the total system. For FFFS the
design life is typically 20-30 years depending on the materials used.
A3.2 Maintenance
A good maintenance strategy starts with an integral system design to keep user costs during the
life cycle as low as possible and minimising potential safety concerns for the tunnel users and
the operations personnel. Therefore, maintainability should be part of the specification for FFFS
and should be incorporated into the design process. Regular maintenance of FFFS will be needed
to meet the performance specifications as well as the reliability and availability demands during
the design life of the system. This is done by periodic inspection, testing and maintenance in
accordance with the manufacturers instructions, and as prescribed in the various national
standards if they exist (e.g. recommendations for the frequency of inspection and testing can in
standards such as NFPA 750 [21] or EN 12845 [66].
System components should be regularly inspected and tested to verify that they function as
intended. Depending on the system design and location of components, this can lead to some
unavailability of the tunnel for components that need visual inspection, maintenance or testing.
There will generally be significant constraints when the systems in a road tunnel under operation
can be maintained. Typically, tunnel maintenance closures take place at night to minimise
disruption to traffic. Therefore, maintenance during these hours is only possible for components
which are installed inside the traffic envelope. In general, wherever possible, the components of
the FFFS should be installed outside the main traffic tube (e.g. section valves in cross passages),
or installed in such a way that they minimise traffic disruptions during maintenance activities
(e.g. the water main suspended on the wall above one lane so the entire tunnel cross section is
not affected). Whenever possible, FFFS components inside the tunnels should be capable of
remote testing so that disruption to the traffic is not necessary or is minimal. Given the range of
systems installed in most road tunnels, maintenance access for FFFS may depend on what needs
to be done for other tunnel systems and elements. Consequently, maintenance access for FFFS
69
may influence the design so that the installed system can be maintained to provide a high
reliability combined with a high availability.
Parts of FFFS that typically require maintenance are listed below. Note that the list is indicative
only and may vary depending on the type of system installed (i.e. Deluge System or Water Mist
System). Also, some items listed below may not be incorporated in every system, like additives
(e.g. foam) or protection against freezing:
power supply to the various components;

water supply;
additives;
pumps;
protection against freezing (heating, circulation);
piping system and supports;
motorised and control valves (section valves);
SCADA system including operational control from the control room;
nozzles;
drainage; and
fuel, lubricant, consumables (e.g. in the case of diesel-driven pumps).
General maintenance frequencies for some components are provided next page. As above, the
list is indicative only and may vary depending on the type of system installed and by manufacturer.
Maintenance after a fire:
all yearly maintenance on system parts affected by the fire;

performance test for pressure and flow; and
replacing or cleaning of nozzles, pipes, brackets, etc. affected by the fire.
If the FFFS are to undergo maintenance or are partially not available, relevant stakeholders such
as the fire brigade and alarm receiving facility should be informed prior to the FFFS being
unavailable. They also should be informed when the FFFS are returned to service.
Special attention should be given to maintenance aspects that could affect the performance of the
system such as change in maintenance contractor, process or material changes, structural changes
and other relevant changes to the system. These changes should be incorporated in the change
management and safety management system of the tunnel owner.
A3.3 Return to service
Before FFFS are returned to service (after regular maintenance, a component failure and
replacement, or a fire event) they should be tested to ensure that they are working properly. This
can be verified by comparing the specific test results with the original acceptance tests and the
previous test results. The extent of verification testing required may vary depending on the
reasons why the FFFS were unavailable.
70
2016R03EN
TABLE 9 - MAINTENANCE ACTIVITIES AND FREQUENCIES

Frequency Activity
Weekly Visual inspection of pump unit on irregularities, leakages, etc.
Pump:
Pump test (start up but without pressurising the main pipes);
Inspection of water tank, pipes and brackets;
Inspection of water filters;
Where combustion engine driven pumps are installed, check level of fuel tank, motor oil, etc.;
Draining condensation pipe in exhaust gas system;
Monthly Check starter battery status for combustion engine driven pumps;
Test pumps additives (e.g. foam);
Test anti-freezing circulation pump;
Test (anti-freezing) trace heating system;
Operational test section valves;
Test nozzles (only if an active component); and
Visual inspection of main pipe work for mechanical damage and other abnormalities.
Testing of batteries;
Inspect batteries, control panels, interface equipment;
Manually operate pressure release valves;
Inspect initialising devices and detectors;
Bi-annually
Inspect pump filters and clean/replace;
Inspect filters anti-freezing pump and clean/replace;
Test section valves; and
Sample check on spare parts that should be in stock.
Visual inspection of nozzle heads and active spraying in all sections (depending on the system and particularly for long tunnels, one can also choose to only
inspect a certain percentage of the nozzles at one time, or at a different frequency e.g. to complete an inspection of all nozzles bi-annually instead of annually);
Check water supply quality;
Change oil pump/diesel engine;
Inspect zone control valves, clean protection boxes;
Annually
Software tests (pump installation);
Software and interface test control;
Test sensor components (level indicators, pressure, temperature);
Check main pumps (including frost protection) according to manufacturers specifications; and
Check all reservoirs for pollution and water quality.
Checking joints/connections;
5 yearly Replacing all high and low pressure hose connections;
Pressure testing of the pipe work (above normal operating pressure); and
Analysing and (if necessary) replacement of additives.
5-12 years Hydrostatic test pressure cylinders (only water mist).
71
Depending on the regulatory requirements in various countries full system tests may be mandated
at regular intervals. Notwithstanding regulatory requirements, a yearly full functional test could
be done during night-time maintenance closures. This means the total system would be tested
from detection to discharge. Tests could include visual tests of the spray behaviour inside the
tunnel or removal and detailed testing of the spray pattern of single nozzles for a certain
percentage (e.g. 25% in Holland) of the sections equipped with FFFS, or a removal and external
inspection of the spray pattern of a percentage of nozzles (e.g. at least 0.5% of all nozzles in
Netherlands). The flow and pressure at the most hydraulically demanding location of the test
sections should be recorded.
APPENDIX 4 - RESEARCH AND EVALUATION PROGRAMS
A4.1 Test programs
Except for Australia and Japan, FFFS started to have greater recognition after full scale testing
was undertaken after several major tunnel fire incidents. These full scale tests showed that FFFS:
limit the peak fire heat release rate;

prevent the spread of fire from vehicle to vehicle; and
provide a certain level of structural protection.
In addition, the fire incident in the Burnley Tunnel in Australia in March 2007 showed that a
Deluge System was effective in allowing the re-opening of the tunnel the next day. This limited
the time the facility was closed and hence any potential losses in revenue stream from tolls.
There have been numerous full scale fire tests in recent years as a result of the changing
international landscape. Tests of FFFS have a wide range of applications due to the objectives of
various entities that perform the tests and the different types of systems (Water Mist or Deluge
Systems). Since the use of FFFS for most of the world is a recent shift in thinking and acceptance
in the industry, the last 10-15 years has provided a wide range of test programs. A list of these
programs follows.
Test programs are sponsored by multiple agencies and frequently do not have similar objectives.
Some testing is sponsored by government entities while other test programs are performed by
private industry. Some test programs are extended installation tests performed under the direction
of the tunnel owner/operator. These separate entities have varying motives for test performance
and each test should be reviewed on its individual merits.
Regardless, certain commonalities can be extracted upon reviewing the test programs available.
Repeatability of results and similarities in findings are an important take-away from various test
programs. This section attempts to consolidate the significant findings of the various test
programs.
72
2016R03EN
TABLE 10 - LIST OF TEST PROGRAMS

WATER APPLICATION RATE
TEST PROGRAM YEAR LOCATION TYPE OF FFFS FUEL
(mm/min unless noted otherwise)
Sweden T-Rex [63] 2013 Runehamar, Norway Deluge 10 Wood pallets
Singapore Test Programs [43] 2012 San Pedro de Anes, Spain Deluge 8 & 12 Wood & plastic pallets
University of Carleton [54] 2009 National Research Council Canada Deluge Varied; 3, 6, 9, 12 & 14 Propane
A73 Tests (water mist) [12] 2008 Runehamar, Norway Water mist Pallets and pool fires
Diesel pools, wood & plastic
M30, Madrid Test Program [46] [47] 2006 San Pedro de Anes, Spain High pressure water mist 0.5-0.72 l/min/m3
pallets
Pool fire; wood pallet & wood
Intermediate Scale Campaign [42] 2006 CSTB, Paris High pressure water mist 0.4-0.8 l/min/m3
cribs
A86, Paris Test Program [62] 2005 VSH, Switzerland High pressure water mist 0.47 l/min/m3 Automobiles
Scale Tests [52] 2005 SP, Boras, Sweden Deluge 5-15 equivalent Wood cribs
A73 Tests (CAF) 2005 Runehamar, Norway CAF NA Wood pallets
1400 l/min for sprinkler;
Sprinkler water curtain, high
UPTUN [50] 2004 DMT, Germany 8 for low pressure Pool fire & wood pallets
pressure water mist
(150 l/min/m3)
Low and high pressure water 1.1-3.2 for low pressure
UPTUN [59] 2004 IF Tunnel, Norway Pool fire & wood pallets
mist 0.5-2.3 for high pressure
A86 Test Program [48] 2003 VSH, Switzerland High pressure water mist 0.33 0.35 l/min/m3 Automobiles
Vehicles loaded with wood
2nd Benelux Tests [45] 2001 Benelux Tunnel, The Netherlands Deluge 12.5
pallets
Pool fire, bus & passenger
Shimizu Tests [53, 56, 58, 60] 2001 Japan Deluge 6
vehicles
Memorial Tunnel Fire Ventilation Test Memorial Tunnel, West Virginia,
1995 Deluge w/AFFF 8 l/s; water application rate not recorded #2 diesel pool fire
Program [41] USA
PWRI Tests [51] 1980 Japan Deluge 6 Bus & passenger vehicles
Offenegg [49, 54] 1965 Switzerland Deluge Not recorded Pool fires
Tests without fully published data
SOLIT2 [24] 2012 San Pedro de Anes, Spain High pressure water mist Not published Wood pallets & diesel pool fire
Low and high pressure water
Mont Blanc Tests [confidential] 2012 San Pedro de Anes, Spain Unknown HGV mock-up
mist and deluge
SOLIT1 [61] 2006 San Pedro de Anes, Spain High pressure water mist Not published Wood pallets
Euro Tunnel Tests [confidential] 2010 San Pedro de Anes, Spain High pressure water mist 4.4 Wood pallets
SINTEF Runehamar Tests [71] 2007 Runehamar Tunnel Water mist Not published Pallets/pool fire
A4.2 Significant findings

73
A4.2.1 FFFS Prevents the Spread of Fire from One Target to Another
Various test programs have established fuel targets as potential combustibles. These targets take
various forms. For instance:
2nd Benelux fire tests (2001): Target was a full-scale tanker truck which was instrumented to
determine heat flux and temperature;
A86 Test Program (2003): A passenger car vehicle abutted up against the burning vehicle;
University of Carleton (2009): A radiation measurement gauge was used to assess the fire
spread; and
SOLIT2 (2012): A pile of wood pallets was placed downstream of the fire.
The emphasis has been to determine whether a fire will spread from vehicle to target. In all tests,
the target fuel did not ignite or would not ignite based on the measured heat flux. FFFS are very
successful in preventing the spread of fires from one fuel source to another.
A4.2.2 FFFS Destroys a Stratified Smoke Layer

Full scale tests have shown that stratification in a smoke layer will be destroyed in the region of
FFFS activation. Every test that has been performed with overhead nozzles has proven this
concept. This destratification is usually constrained to the area where the system is applied.
It should be noted that a smoke layer naturally de-stratifies some distance from the fire. Fast
growing fires may have destratified smoke typically around 3 minutes after smoke propagation.
This phenomenon has been noted in numerous tests and extensively recorded in the Memorial
Tunnel Fire Ventilation Test Program [41].
A4.2.3 FFFS Reduce Visibility within the Zone where Activated, Even Without Fire
Visibility in the region of activation of FFFS is significantly reduced even without a fire event.
Test programs have repeatedly shown this effect. Visibility is affected as if there is a heavy
downpour of rain.
A4.2.4 Radiation Effects from a Fire are Reduced

Radiation is a key contributor to flame spread, endangers tunnel occupants and also prevents
fire-fighter access to the immediate region around a fire. Many full scale test programs have
measured this phenomenon and have recorded significant reductions in radiation flux with the
activation of FFFS.
2nd Benelux fire tests (2001);

A86 Test Program (2003);
University of Carleton (2009); and
SOLIT2 (2012).
A4.2.5 P eak Temperatures are Reduced and the Region of Tunnel Impacted by High Heat
Effects is Significantly Minimised
FFFS have shown the capacity to reduce peak temperatures directly above the fire, but more
significantly, the region of high temperature impact in the tunnel is greatly reduced.
In the University of Carleton work, temperature effects were reduced significantly for a 5 and
10 MW fire HRR. Measurements in SOLIT2 have shown that the temperature of a potential
74
100MW fire below the tunnel ceiling (0.2 m below ceiling) drops significantly at a location
10 m behind the fire load when the water mist system is started.
A4.2.6 Fire Heat Release Rate can be Reduced

A fire can be extinguished, but it depends on the water application rate, the water distribution
and the activation time of the system. If a fire is not extinguished with application of FFFS,
it may continue to burn but at reduced HRR [23].
A4.2.7 Steam Generation is not Sufficient Enough to be Considered a Threat

One of the key conclusions from the Offenegg tests in 1965 was that steam generation was a
viable threat to life. One of the objectives of the 2nd Benelux Tests was to review this position.
Using a deluge system with a 12.5 mm/min flow rate, steam generation was not shown to exist
in appreciable quantities. Future fire tests also did not show steam generation, though these were
not stated objectives of these tests.
A4.3 Summary
Various broad aspects of FFFS based on large scale testing have been reviewed. These are
discussed in detail and conclusions given for each aspect. In general, two distinct water application
rates were used during full scale testing. Deluge Systems tests normally used a water flow rate
of 10 to 12.5 mm/min. High pressure Water Mist Systems normally used a water flow rate of
around 4 mm/min. The principal mechanisms of fire suppression using these two types of
systems are different. A Deluge System prevents fire growth and reduces fire heat release rate
mainly by fuel surface cooling. A Water Mist System mitigates a fire mainly by gas cooling.
The technical benefits of FFFS have been strongly demonstrated via full scale testing. FFFS
clearly provide an effective means of managing a fire incident remotely. While FFFS have not
been shown to reliably quench or extinguish a fire, the evidence shows that FFFS effectively
contain a fire incident and grant the local fire brigade the time and opportunity to approach a fire
incident to address it directly. In addition, the region around a fire incident that may be influenced
by high temperatures is significantly reduced to the fire location.
Visibility is reduced significantly by the action of FFFS. This may impair egress, but the reduced
fire HRR and growth rate also provides additional time for the fleeing motorist who may be wet,
but will not be endangered by the tunnel environment.
In most of the full scale tests, the fires were neither extinguished nor fully suppressed, instead
conditions around the fire were controlled. Therefore, there is a growing consensus that FFFS
only mitigate the fire effects in tunnels. To successfully extinguish a fire, the performance of
FFFS needs to be improved.
There are also some arguments that the performance of a Water Mist System is better than a
Deluge System. However, under the tested water flow rates, the performance of the Deluge
Systems is better on controlling the size of the fire, whereas the Water Mist System is better on
thermal management of the environment. Further, it should always be kept in mind that the
Deluge Systems and Water Mist Systems discussed here use significantly different water flow
rates and droplet characteristics.
75
APPENDIX 5 - MODELLING
A5.1 Introduction
Computational Fluid Dynamics (CFD) has been used to accurately model many aspects of tunnel
fire safety for more than 20 years. It can be described as a mature tool to gain better insight into
smoke movement and assess tenability conditions during a fire.
To install FFFS in a given tunnel, the present state of the art implies performing full-scale fire
tests to assess the optimum parameters and effectiveness of FFFS in the event of a fire.
However, CFD has great potential for long-term usefulness and might be considered as a valid
supplement or even an alternative in the future. This section addresses the essential issue of
using CFD for FFFS in tunnels, noting that with current active research this is subject to
change.
A5.2 CFD framework for FFFS

As FFFS are discharged inside a tunnel, water droplets mix with the gas flow. The fluid flow in
the spraying zone contains therefore two physical phases, the gas phase (e.g. air, smoke) and a
liquid phase, the water droplets. To fully describe the state of each phase at any point in the
tunnel, two different sets of conservation equations are needed.
For tunnel fire computations, the gas flow is typically described with an Eulerian reference
frame methodology and solved through Reynolds Averaged Navier-Stokes (RANS) or Large
Eddy Simulation (LES) numerical methods. To account for water droplets interaction with the
fire conditions, these equations need to be modified to include inter-phase mass, momentum and
energy transfer. Various treatments of the dispersed phase can be employed but Lagrangian
particle tracking is conventionally used. Individual particle trajectories are computed with
equations based on the moving particle location. For a large number of particles, computational
parcels can be used where each parcel represents a cloud of many particles with the same
characteristics.
A5.3 Input data

For specific FFFS, manufacturers usually provide the operation pressure and the water flow
rate, sometimes expressed as an application rate per unit surface area or volume. When CFD
modelling of FFFS is involved, these characteristic parameters are not enough, as the
computation requires initial conditions for the particle trajectories. The two main input data are
therefore the droplet size distribution and the discharge velocity at the nozzle. The latter can be
derived from the operation pressure at the nozzle if the individual nozzle K-factor is provided.
The former can only be determined through individual nozzle characterization experiments at
laboratory scale.
For a specific nozzle, size distributions are measured at different heights and angles to derive a
representative size distribution and a cone angle for the nozzle spray. Numerous optical
measurement methods can be used to determine droplet velocity and/or size, but only the Phase
Doppler Anemometry (PDA) method is able to measure both at the same time. The method is
based on light scattering interferometry and the Doppler effect, and requires a large laser source
and multiple photo detectors to assess droplet size and velocity at a given point. The data obtained
are processed to provide statistical parameters than can be input into CFD models.
76
A5.4 Validation
There are a number of definitions of validation and one that is widely accepted defines validation
as the process of determining the degree to which a model is an accurate representation of the
real world from the perspective of the intended uses of the model.
This definition implies that:
validation involves a process, not a single test or comparison;

it requires an assessment of accuracy, signifying that error and uncertainty are important; and
the context is important, in that different intended applications of the model may have different
requirements.
There have been several full-scale test programs that have been used to perform comparisons of
various aspects for CFD modelling. The challenges are in the areas of pyrolysis and combustion
modelling, and the spray modelling including wall impingement. The phenomenon of pyrolysis
is very complex, however, approximate models can be applied in some scenarios. This part
corresponds to the highest uncertainty in CFD modelling of FFFS. Clearly, obtaining accurate
solutions is much more challenging than gas-phase calculations for tunnels, not just because the
gas and liquid phases must be treated separately but, in addition a number of sub-models such as
those accounting for inter-phase mass, momentum and heat transfer have to be carefully selected
and validated.
CFD models used for assessing the effect of FFFS on a tunnel fire were developed to handle a
number of multiphase flow situations, including gas-liquid, liquid-solid, gas-solid or any
combination of the three phases. They may have been verified or even validated for a given
context but not for tunnels specifically. Therefore, predictions of the models need to be compared
with experimental data from full-scale tests or scaled ones where the scale factor is large enough
to have little influence over the key physical phenomena.
The experimental results should thus include relevant physical data, which are of interest to CFD
validation, and should capture initial/boundary condition used in the model. A number of
different variables in various locations need to be measured. They include HRR, gas temperature,
heat flux, gas velocity and concentrations.
A5.5 Limitations
CFD simulations of FFFS become computationally intensive as the number of Lagrangian
particles increases. Only transient calculations for the two phases can be performed to properly
capture droplet behaviour. The correlations used in inter-phase source terms (e.g. drag
relationships) are empirical or semi-empirical. Selecting the models that give an accurate
representation of physical reality requires experience on the part of the CFD modeller.
The area that has the most uncertainty in the CFD modelling of water/fire interaction relates to
combustion models. Models have difficulty predicting the interaction of droplets on the fuel
due to the typical size of a grid cell versus the much smaller scale of the interactions that the
model is attempting to predict. Combustion is fairly well understood, but it is difficult to model
the full complexity, thus simpler combustion models are generally applied to road tunnel
applications.
77
Capturing the suppression of a fire using CFD is an evolving field due to the difficulty in
modelling the micro-scale interactions among the fuel, oxidant and suppressant. Fire models that
use volumetric heat sources have limited provision for modelling flame water interaction as a
gross scale must be applied. Manually reducing heat release rate can be accomplished, but this
process requires a comprehensive database to develop an empirical relationship between the
HRR, and water droplet size and quantity. If the HRR and combustion product yield are well
understood a priori, CFD models are able to capture far field effects well (temperature, gas
concentration, heat fluxes) [64].
Other approaches to CFD models of FFFS use more detailed combustion models. With the
mixture fraction method, a quantity of vaporised fuel is injected into the computation domain
from a burning surface. Combustion occurs when the fuel vapour comes into contact with air
and sufficient heat. This approach incorporates the physics of a flame, HRR and combustion
product yield in more detail than a volumetric approach. The quantity of fuel released from the
burning surface in the CFD model can be determined via pyrolysis models or a priori specified
heat release rates and fuel parameters. When FFFS are included their effects can be modelled
within the pyrolysis model or with an empirical approach. In the pyrolysis model approach,
water spray reaches the burning surface and cools it, and the spray also interrupts heat transfer
to the surface, thereby slowing fuel pyrolysis and limiting the fire HRR. As for the empirical
approach, the amount of heat released from a burning surface is reduced in proportion to the
quantity of water reaching the surface [65].
However, both methods rely on empirical parameters. One major challenge with pyrolysis
modelling is the difficulty in obtaining relevant material properties. With the empirical approach
experiments are conducted to find the relationship between water reaching the burning surface
and fire heat release rate, however, this approach models a micro-scale phenomenon using
macro-scale observations. Both methods are affected by the wide range of possible fire loads in
a tunnel, which have various geometry and fuels. For these reasons, these methods are in their
developmental stages and are not typically used in current project design to determine the design
fire HRR. There is active and ongoing research to better understand and model the impact of
FFFS on the fire [65].
In summary, CFD models with FFFS and a prescribed HRR can be used with a high degree of
confidence to predict temperatures, radiative heat flux and smoke behaviour in regions remote
from the immediate fire. Methods exist to predict the interaction of FFFS with the HRR.
However, these methods involve more complex physics, a greater range of length scales, and
they are influenced by uncertainty in the actual fire geometry. As such, the prediction of HRR
and combustion products yield using CFD is an evolving area of practice.
APPENDIX 6 - SUSTAINABILITY
Sustainability involves broad concerns about economic, social and environmental objectives.
For transport infrastructure, construction and operation increasingly include measures to
improve sustainability by reducing environmental impact over the operational lifecycle. Such
impacts are typically assessed in terms of carbon emissions. Most efforts to date to assess carbon
emissions from facilities focus on normal operating conditions, but some consideration has been
given to the influence of fires and FFFS.
78
To assess the impact of FFFS, it is important to understand the overall context of carbon
emissions. The total carbon emission over the lifecycle of a road tunnel will include the sum of
emissions from:
construction (including materials, transportation and equipment usage);

normal operations and maintenance (notably power consumption);
periodic refurbishments (equipment and transportation for disposal of redundant materials, and
further new materials, transportation and equipment usage); and
tunnel fires (and other incidents), mitigation responses and subsequent repairs.
The benefit of FFFS will be to reduce the fire risks and corresponding carbon emissions. This
will be partly negated by the carbon emissions associated with FFFS themselves. To quantify the
impact of FFFS, it is therefore necessary to quantify the potential impact of fire risks on the
overall carbon emissions over the lifetime of the tunnel. A basic framework for such calculations
is illustrated below (adapted from Gritzo [7]).
The total carbon emission (TCE) can be expressed by:
TCE = CEconst + CErefurb + LCEo+m + LCEf
Carbon emissions from construction, CEconst and periodic refurbishment, CErefurb can be
considered to be discrete events. These emissions are typically referred to as embodied emissions
given their inclusion in the physical elements of the tunnel rather than resulting from tunnels
operations.
Emissions from operations and maintenance, LCEo+m can be considered on an annual basis and
included in life cycle analysis by multiplying annual rates of emission, ACEo+m by the lifetime of
the tunnel:
LCEo+m = Tlife ACEo+m
ACEo+m represents the annual rate of emission for operations and maintenance and is often
referred to as the carbon footprint. Power consumption for tunnel lighting is often an important
component of this footprint. The carbon footprint for a fire suppression system would be expected
to be relatively small, given its infrequent activation.
LCEf represents carbon emissions associated with the fire risks over the lifetime of the tunnel.
These include direct emissions (of carbon dioxide and soot, for example) as well as indirect
emissions due to asset damage and release of embodied emissions, both from the tunnel assets
and from the goods carried by the vehicles involved in the fire. This can be expressed by:
LCEf = Tlife ff (mf ECO2 + Fr CEemb)
where:
f f annual frequency (fires/year)

m f mass of material burned (kg fuel)
ECO2 CO2 released per unit material burned (kg CO2/kg fuel)
79
Fr fraction of material to be replaced during repairs (-)

CEemb total embodied CO2 emissions (kg CO2/kg) [= CEconst + CErefurb]
Reflecting these different contributions, the impact of fire on the lifecycle carbon emissions is
illustrated in illustration 9. The impact of FFFS can be measured by the reduction of the fire risk
term.
Illustration 9 - Lifecycle carbon emissions

Conceptual Diagram of Contribution of Risk Factors for Lifecycle Carbon Emissions,
Gritzo L.A., Doerr W., Bill R., Ali H., Nong S. and Krasner L.,
The Influence of Risk Factors on Sustainable Development FM Global Research Division
Note that illustration9 is not to scale. The relative magnitudes of carbon emissions for the
different contributions vary widely between tunnels and therefore generic values are not
suggested here. The estimation of carbon emissions requires a range of information relating to
construction and refurbishment, operations and maintenance. There are various sources of
construction industry information, including databases of estimated CO2 emission values for
different construction activities. A tailored approach is likely to be required for a road tunnel.
When dealing with buildings, carbon emission parameters are typically expressed per unit area,
but this would be inappropriate for road tunnels due to their linear nature and the fact that fire
risks are principally associated with vehicle traffic.
Another consideration is that a tunnel and its vehicular traffic represent a broad inventory of
materials, potential fire loads and fire risks. The risk profile typically encompasses a broad
range of potential scenarios (consequences and probabilities) rather than the single set of
representative values implied above.
To summarise, the sustainability implications of FFFS can be evaluated in terms of carbon

emissions. The benefit of tunnel FFFS will be to reduce the fire risks and corresponding carbon
emissions. The significance of this reduction depends on the overall carbon emissions over the
lifetime of the tunnel.
Copyright by the World Road Association. All rights reserved.
World Road Association (PIARC)

Tour Pascal - 19e tage
92055 La Dfense cedex, FRANCE
International Standard Book Number 978-2-84060-375-7

Frontcover Marina Coastal Expressway, Singapore

Fixed Fire Fighting in Road Tunnels Current Practices and Recommandations

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fixed Fire Fighting in Road Tunnels Current Practices and Recommandations

Uploaded by

Copyright:

Available Formats

www.piarc.

FIXED FIRE FIGHTING

Copyright by the World Road Association. All rights reserved.

World Road Association (PIARC)

International Standard Book Number 978-2-84060-375-7

The contributors to the preparation of this report are:

Radim Bajger (Czech Republic) Norris Harvey (USA)

FIXED FIRE FIGHTING SYSTEMS

1.1. PURPOSE OF THIS REPORT..........................................................................................................................3

2.1. PIARC PUBLICATIONS..................................................................................................................................6

4. DESIGN CONSIDERATIONS ................................................................................................................................... 17

4.1. RISK ASSESSMENT.......................................................................................................................................18

5. SYSTEM DEFINITION / PROCUREMENT..........................................................................................................28

5.1. SYSTEM DEFINITION..................................................................................................................................28

6. RESEARCH AND ANALYSIS...................................................................................................................................34

6.1. RESEARCH PROGRAMS..............................................................................................................................34

7. CONCLUSIONS AND RECOMMENDATIONS....................................................................................................36

APPENDIX 1 - QUESTIONNAIRE DATA..........................................................................................................44

1.1. PURPOSE OF THIS REPORT

1.2. SCOPE OF THE REPORT

1.3. TARGET GROUP

1.4. THE ROLE OF FIXED FIRE FIGHTING SYSTEMS

provide early suppression and control of a fire event;

1.4.1. Deluge Systems

Deluge Systems are typified by a zoned water application, characterised by a significant

1.4.2. Water Mist Systems

1.4.3. Common Features and Variances

Both systems are characterised by:

use higher pressures than Deluge Systems;

2.1. PIARC PUBLICATIONS

2.1.1. 1999 - Fire and Smoke Control in Road Tunnels

2.1.3. 2008 Road Tunnels: An Assessment of Fixed Fire Fighting Systems

2.2. OTHER REPORTS AND GUIDANCE

2.2.2. Other Relevant Reports

compliance with local regulations and guidelines, including legal considerations;

Design considerations are addressed in chapter 4, page17.

3.2. COMPLIANCE WITH REGULATIONS, GUIDELINES AND LEGAL

3.3. RISK ASSESSMENT

3.4. LIFE SAFETY

3.5. ASSET PROTECTION

FFFS may be considered as a compensatory measure in fire-engineering design. It is recognised

3.6. ADDITIONAL TRAFFIC REGIMES

3.7. FIRE-FIGHTING RESPONSE

3.8. OPERATIONS, MAINTENANCE AND REGULAR TESTING

Maintenance of FFFS includes all technical, operational and administrative requirements to

Further details on testing and maintenance are given in appendix 3, page68.

3.9. COST BENEFIT CONSIDERATIONS

construction (including materials, transportation and equipment usage);

3.11. LEGAL CONSIDERATIONS

water suppression characteristics (mist or deluge);

design fire HRR and fire growth characteristics;

4.1. RISK ASSESSMENT

4.2. TYPE OF SYSTEM

Illustration 1 Deluge electric pump and panel

4.3. WATER DISCHARGE CHARACTERISTICS (MIST OR DELUGE)

Illustration 2 Deluge discharge

4.4. WATER SUPPLY

Illustration 3 Deluge water storage tank

4.6. SPACE CONSIDERATIONS

4.7. FIRE DETECTION AND ACTIVATION STRATEGY

Activation of FFFS can be:

4.8. TUNNEL ENVIRONMENT

4.9. SYSTEM INTEGRATION

TABLE 1 - FACTORS ASSOCIATED WITH SYSTEM INTEGRATION