ITA WG6 - 2017 FINAL Edition

Structural Fire Protection For Road Tunnels
ITA Working Group 6

Maintenance and Repair
N° ISBN: 978-2-9701122–0-4 ITA REPORT n°18 / april 2017
21611-AITES-REPORT-18-WG6.indd 1 02/06/17 18:33

ITA Report n°18 - Structural fire protection for road tunnels - N°ISBN: 978-2-9701122–0-4 / April 2017
Layout : Longrine – Avignon – France – www.longrine.fr
The International Tunnelling and Underground Space Association/Association Internationale des Tunnels et de l’Espace Souterrain (ITA/AITES) publishes this report to, in accordance with its
statutes, facilitate the exchange of information, in order: to encourage planning of the subsurface for the benefit of the public, environment and sustainable development to promote advances
in planning, design, construction, maintenance and safety of tunnels and underground space, by bringing together information thereon and by studying questions related thereto. This report
has been prepared by professionals with expertise within the actual subjects. The opinions and statements are based on sources believed to be reliable and in good faith. However, ITA/AITES
accepts no responsibility or liability whatsoever with regard to the material published in this report. This material is: information of a general nature only which is not intended to address the specific
circumstances of any particular individual or entity; not necessarily comprehensive, complete, accurate or up to date; This material is not professional or legal advice (if you need specific advice,
you should always consult a suitably qualified professional).

ITA Working Group 6

Maintenance and Repair

>> Table of contents
1. introduction.......................................................................................................................................6
1.1 background...................................................................................................................................6
1.2 scope................................................................................................................................................6
1.3 reasons for developping guidelines....................................................................................6
1.4 document authors......................................................................................................................6
2. design criteria for fire resistance.........................................................................................7
2.1 introduction..................................................................................................................................7
2.2 DATA ON TUNNEL FIRES...................................................................................................................7
2.3 DESIGN CRITERIA ESTABLISHED BY PIARC.....................................................................................9
2.3.1 PIARC Recommendations..........................................................................................................9
2.3.2 Pending questions....................................................................................................................10
2.4 ITA CLASSIFICATION........................................................................................................................11
2.5 TIME-TEMPERATURE CURVE VERSUS MEGAWATTS....................................................................12
2.5.1 Issues with conversion.............................................................................................................12
2.5.2 Concerns about megawatt output of bus fires.........................................................................12
2.5.3 Recent research and modelling................................................................................................12
3. LINING MATERIAL BEHAVIOUR...........................................................................................................14
3.1 CONCRETE.......................................................................................................................................14
3.1.1 Heating.....................................................................................................................................14
3.1.2 Thermal properties...................................................................................................................14
3.1.3 Mechanical properties..............................................................................................................15
a/ if no temporary damage of the concrete structure is allowed........................................................16
b/ if temporary damage of the concrete structure is allowed.............................................................17
3.1.4 Spalling.....................................................................................................................................17
3.2 REINFORCEMENT..............................................................................................................................18
3.3 PROTECTIVE MEASURES..................................................................................................................19
3.3.1 Commonly used product types................................................................................................19
3.3.1 Non-suitable fire protection for use in tunnels...........................................................................19
3.3.2.1 Paint products and systems..................................................................................................19
3.3.2.2 Magnesium Oxy-Chloride based materials............................................................................19
a/ Corrosive effects of leaching Chloride ions....................................................................................19
b/ Impact of Hydrochloric acid (HCl) on tenable environment............................................................20
c/ Independent research conclusions................................................................................................20
4. TUNNEL CLASSIFICATION...................................................................................................................21
4.1 GENERAL...........................................................................................................................................21
4.2 TUNNEL CLASSIFICATION DESCRIPTION.......................................................................................21
4.3 TUNNEL CLASSIFICATION FOR RISK ANALYSIS.............................................................................21
4.4 TUNNEL CLASSIFICATION SKETCHES............................................................................................21
5. STRUCTURAL ELEMENTS...................................................................................................................27
5.1 GENERAL...........................................................................................................................................27
5.2 STRUCTURAL ELEMENTS.................................................................................................................27
5.3 OBJECTIVES AND FIRE RESISTANCE LEVELS.................................................................................27
4 ITA Working Group 6 Maintenance and Repair

>> Table of contents
5.4 FIRE RESISTANCE LEVELS................................................................................................................27

5.5 EVIDENCE OF FIRE RESISTANCE OF PROTECTED STRUCTURES................................................28
5.6 MINIMUM REQUIREMENTS FOR MAIN STRUCTURES.....................................................................28
5.7 LINING, MATERIAL TYPES AND FIRE BEHAVIOUR...........................................................................28
5.7.1 Loads.......................................................................................................................................29
5.7.2 Segmental capacity..................................................................................................................30
5.7.3 Capacity of elements in bending...............................................................................................31
5.7.4 Fire resistance..........................................................................................................................31
5.8 LININGS..............................................................................................................................................31
5.8.1 Concrete linings........................................................................................................................31
5.8.2 Steel linings..............................................................................................................................32
5.8.3 Cast Iron linings........................................................................................................................32
5.8.4 Masonry linings.........................................................................................................................32
5.8.5 Gaskets....................................................................................................................................32
5.8.6 Packers....................................................................................................................................33
5.8.7 Immersed tunnel joints.............................................................................................................33
5.9 SUSPENDED CEILINGS.....................................................................................................................33
5.9.1 Concrete ceilings......................................................................................................................33
5.9.2 Steel ceilings.............................................................................................................................33
5.9.3 Multiple sides fire exposures.....................................................................................................33
5.9.4 Loads and structural behaviour................................................................................................33
5.10 SUPPORTED FLOORS AND DECKS...............................................................................................34
5.11 ANCHORAGES.................................................................................................................................34
5.12 MITIGATION TECHNOLOGY FOR TUNNEL STRUCTURES.............................................................34
5.12.1 Bored Tunnel (Horse Shoe Tunnel) in stable rock conditions...................................................34
5.12.2 Example of Circular Bored Tunnel (Shield Tunnel)...................................................................34
5.12.3 Mitigation measures for protection of ceilings in the Liefkenshoek Road tunnel......................35
5.12.4 Mitigation measures for protection of anchorages in traffic space..........................................35
5.12.5 Protection of structural elements............................................................................................36
5.12.6 Type of fire protection materials..............................................................................................37
5.12.7 Development of fire resistant concrete....................................................................................38
5.13 CURRENT PRODUCTION LISTS OF FIRE RESISTANT MATERIALS................................................38
6. SUMMARY AND RECOMMENDATIONS................................................................................................39
7. BIBLIOGRAPHY........................................................................................................................................41
7.1 REFERENCES FOR SECTION 1.........................................................................................................41
7.5 SOME REFERENCES TO GO FURTHER............................................................................................42
APPENDIX 1 - ROAD TUNNEL FIRE HISTORY..........................................................................................44
APPENDIX 2 - LI AND INGASON MODEL..................................................................................................46
APPENDIX 3 - RUNEHAMAR BROCHURE................................................................................................47
Structural Fire Protection For Road Tunnels 5

2 >> S
1 Introduction
afety requirements in all stages of construction
1.1 Background This document is intended to be a guideline The primary contributors for this document
and is to be used for road tunnels only are:
Fire resistance of tunnel structures is and not for rail, mass transit, or pedestrian
an important issue. If it is not properly tunnels. However the basic principles for Mr. Henry Russell U S A
addressed, fire in a tunnel can result in loss the protection of tunnels and underground (Animateur/Editor, 2002 - Present)
of life to both tunnel users and the fire and structures may be applied to other types Mr. René van den Bosch Netherlands
rescue services. Resulting economic losses of structures; in such cases special (Vice Animateur, 2005 - Present)
for both the tunnel owner/operator and to consideration must be given to the particular Mr. Martin Muncke Germany
the local economy and environment can be application and its own unique operational (Vice Animateur, 2003 - 2005)
catastrophic. and other site-specific elements. Prof. Alfred Haack Germany
(ITA Tutor, 2001 - 2004)
This document is produced by Working This document is for informational purposes Prof A Bai Yun China
Group 6 Repair and Maintenance of only and applicable codes, standards and (ITA Tutor 2004 - Present)
Underground Structures and has been local regulations must be consulted for Dr. Kees Both Netherlands
solely developed by the members of compliance to specific structural and life Mr. Jan Jonker Netherlands
that Working Group. The purpose of this safety requirements of the locale in which Mr. Per Boman Sweden
effort is to develop recommendations for the structure is located. Mr. Andreas Henke Switzerland
fire protection for concrete road tunnel Mr. Niels P. Høj Denmark
structures. However, many of the issues 1.3 Reason for Developing Mr. Volker Wetzig Switzerland
addressed in these guidelines are relevant to Guidelines Mr. Gunnar Gjaeringen Sweden
rail tunnels and this will be referenced as part Mr. Paul Sparrow United Kingdom
of this document at a later date. Between 1990 and 2010, there have been a
number of serious underground fires in road The primary contributors wish to express
Full scale fire tests carried out as part of tunnels. These fires have caused extensive their appreciation to the many members
the EC Funded Research Project “UPTUN” loss of life and severe collateral loss to the of the Working Group, who have provided
and privately funded Runehamar Fire Tests infrastructure. Aside from the tragic loss of stimulating debate and comments on this
[1.1] have shown that fires in tunnels can be life, the long-term financial effects to the local document, and also extend their gratitude
much more severe than previously assumed infrastructure, the loss of public confidence in to Mr. Michel Deffayet and Mrs. Bérénice
(even with non-hazardous goods). For this the safe use of tunnels have necessitated the Moreau.
reason, this document makes reference to development of safety recommendations.
the time - temperature curves developed This document is intended for use to
in these tests. It also references the relative identify the categorisation of road tunnels
heat release curves (RHR) developed from and propose methods for the protection
these full-scale fire tests as separate graphs. of the structural elements. The behaviour
of structures is a key factor to allow users
1.2 Scope to evacuate and rescue personnel to enter
the scene and effectively perform their
As an addition to PIARC guidelines, the required duties and to limit damage to the
scope of this ITA document is to provide tunnel. Improved specifications for tunnel fire
recommendations for techniques and resistance are required in order to mitigate
materials to answer these structural the consequences of a serious fire, which
requirements and make tunnels and their could result in structural failure or complete
ancillary structures more resistant to fire collapse.
damage. These recommendations take
into consideration the time - temperature 1.4 Document Authors
curves as recommended by others and
develop suitable means and methods for This document has been developed by the
the protection of the structures. The aim members of ITA working Group 6.
or focus of the protection may vary from
preventing minor damage to preventing a
total collapse both during the fire event and
during the rescue operation.

3
2 >> P
Dlanning
esign criteria
for emergencies
for fire resistance
2.1 Introduction • And the new report Fixed fire fighting Figure 2.1 indicates the temperature
systems in road tunnels: current practices distribution within the test tunnel. There
The design of tunnels to be resistant to and recommendations [2.4] published in seems to be little difference between the
damage as a result of vehicle fire is an 2016 (update of the 2008 version). temperature distributions at the ceiling and
important issue for the construction of new near the road surface, when the cause of
tunnels and the rehabilitation of existing 2.2 Data on Tunnel Fires the fire is Heavy Goods Vehicles (HGV)
underground facilities. The objectives of and public buses. In the cases of plastic
tunnel structural fire resistance associated Section II.4.1 of the PIARC report of 1999 cars and private cars, the temperature is
to safety equipment and smoke control [2.1] indicates the heat release from vehicle relatively low at the road surface when
systems are to allow for users to evacuate fires is dependent on many variable factors compared with that at the ceiling.
safely in the event of a tunnel fire, rescue such as:
operations to be performed under safe • Number of vehicles involved in a fire The reason for this is that HGV and buses
conditions, and tunnel suffers minimal have a relatively large cross-sectional
• Type of vehicles (passenger cars, coaches,
damage. area, or in other words the ratio between
heavy goods vehicles, petrol tankers)
projected area of the vehicle and tunnel
The tunnel system must be protected from • The type and quantity of flammable cross-section (blockage ratio), which
collapse during a specified time period. material available is relatively high. The blockage ratios
The potential for collapse is particularly • Rate and method of extinguishing the fire between vehicle and cross sectional area
important for submerged tunnels (Immersed • Cross section of tunnel structure of actual road tunnels are relatively lower
Tube Tunnels), and tunnels in urban than those for the fire tests. This means
environment that are located under other The temperature is highest on exposed that the measured values in Figure 2.1
buildings or structures. The second focus surfaces and particularly at the higher include some allowance of safety for the
of this document is to ensure appropriate elevations of the tunnel structure. This is correspondence to tunnel safety planning.
protection of property, which also involves illustrated in Figure 2.1 Even if the conditions pertaining to fire
indirect costs associated with the disruption source in the tunnel are similar, temperature
to business, the local economy and the distribution is influenced by the tunnel
restoration of the facility to normal operation. configuration. The main reason for this
is the effects of heat radiation from the
The World Road Association (PIARC) has vehicles as measured by thermcouples,
issued several reports related to tunnel which measure gas temperatures.
fires. These reports have been prepared by
PIARC Working Group 6 Fire and Smoke
Control and edited by the PIARC Technical
Committee on Road Tunnels (now Technical
Committee on Road Tunnel Operation – D5) :
Figure 2.1 : PIARC Maximum Temperatures within
• The report Fire and Smoke Control in
Tunnel Cross Section [2.1]
Road Tunnels [2.1] published in 1999: This
document is the results of the review of PIARC has also established from tests the
numerous fire scenarios and actual case limits of the areas affected by a typical
histories. The report has identified typical tunnel fire. These limits were based on a
scenarios for heat generation caused by a single fire event occurring in a tunnel. This
fire event. information is useful structurally in the
• The report Systems and Equipment for determination of areas to be inspected for
Fire and Smoke Control in Road Tunnels damage. However, since the location of
[2.2] published in 2004: It includes a a fire within a tunnel is random and may
Figure 2.2 : PIARC Maximum temperatures in the ceiling
final recommendation on the question of occur at any location, the entire tunnel area of the tunnel [2.1]
design criteria for resistance to fire for road must be designed to resist fire. Figure
tunnel structures. 2.2 documents the maximum ceiling
temperature in relation to the fire location In addition to the maximum gas temperature
• The report Road tunnels: Operational
for various scenarios as found in the and limits of various fire scenarios, the
strategies for emergency ventilation [2.3]
Eureka tests. report also present temperature vs. time
published in 2011;
duration plots for the various types of fires

2 >> Design criteria for fire resistance
encountered in the Eureka Tests as shown in

Figure 2.3 and the full scale fire tests carried
out in the Runehamar Tunnel in Norway in
September 2003 in figures 2.4 and 2.5.
Temperatures estimated in actual tunnel fires
such as the Mont Blanc and St. Gotthard
should also be considered.
Figure 2.4 : Heat Release from Runehamar tunnel test
Figure 2.3 :Time Dependence Temperature Data from

Eureka Test Program [2.1]
With all the data available measured from

real large scale fire tests with real vehicles
it is possible to identify the maximum ceiling
gas temperatures that can be obtained for
different types of vehicles.
A simplified model was developed by

Ingason and Li [2.5]. To illustrate this model,
two tables are presented below. For further
details on the model, see also Appendix 2.
• Table 2-1 concerns pool-fires, assuming a

tunnel height of 5,5 m;
• Table 2-2 concerns vehicle fires, with a
tunnel height 4,5 m. Figure 2.5 : Temperature recordings from the Runehamar tunnel Tests
Height of tunnel (m) Elevation of fuel-load (m) Area of Fire (m2), assuming a maximum HRR of 1,5 MW/m2 HRR (MW) Maximum temperature (C)
5,5 0 33,3 50,0 634*
5,5 0 50,0 75,0 880*
5,5 0 66,7 100,0 1113*
5,5 0 83,3 125,0 1337*
5,5 0 100,0 150,0 1350
5,5 0 133,3 200,0 1350
Table 2 1: Pool fires

Comments: related to Area

Height of tunnel (m) Elevation of fuel-load (m) Area of Fire (m2) HRR (MW) Maximum temperature (C)
of Fire, and HRR
4,5 1 7,2 5,0 206* 1 car: Fire area = 1,6*4,5 m2
4,5 1 9,0 15,0 558* Minivan: 1,8*5 m2
4,5 1 38,3 35,0 1006* Bus: 2,55*15 ("small"/old)
4,5 1 38,3 50,0 1350 Bus: 2,55*15 (new/ski)
4,5 1 38,3 50,0 1350 LGV: 2,55*15
4,5 1 38,3 100,0 1350 HGV: 2,55*15
Table 2.2 : vehicle fires * Higher if flames touch the wall
2.3 Design Criteria Established

by PIARC
After the publication of its 1999 report, the

PIARC Working Group 6 Fire and Smoke
Control continued its work on resistance to
fire of road tunnel structures. This paragraph
presents its final recommendation on the
question of design criteria for resistance
to fire for road tunnel structures, which
has been included in the 2004 PIARC
report System and Equipment for Fire and
Smoke Control in Road Tunnels [2.2], and
subsequent updates in 2007.
2.3.1 PIARC Recommendations

Figure 2.6 : Recommended Temperature versus Time Curves for ZTV, RWS, HCinc, and ISO Standards [2.6]
A preliminary and basic criterion to be met

by any tunnel structure is that there should of escape and the time conceived for of 400-450°C during the time necessary to
not be any risk of progressive collapse: the evacuation. There should be no collapse fight fire (in a tunnel, such temperatures can
local failure of any element should not lead during this period that can affect the zones produce a radiation level of about 5 kW/m2,
to an increased load on other parts of the where there may be users or rescuers. which is the maximum tolerable value for
structure which may cause their failure. firemen).”
Spalling of the structure can occur in the
There are several time-temperature curves early stages of a fire but no incidents have Shelters should only be provided in a tunnel
proposed to the date. Figure 2.6 sketches been reported where it has had major if there is an escape way for rescuers to
the ISO 834, RWS, RABT (former ZTV) and consequences for firemen, although it reach the users waiting in the shelter and
a modified Hydrocarbon (HC) curve, HCinc, may indicate a rapid deterioration of the assist them to the outside. If such shelters
in which the temperature are multiplied by structure. The main concern at the time are available, then a resistance of about
a factor of 1300/1100 from the basic HC of fire service intervention would be the two hours would be needed for protection
curve of Eurocode 1 Part 2-2. collapse of items, such as jet fans, signs prior to rescue.
or lights from the tunnel ceiling or walls.
Recommendations for design of This question of fire resistance has been The overall duration defined by the curve
the structure should consider the addressed in the PIARC 1999 and 2007 will need to be considered. For instance, in
aforementioned time-temperature curves, reports in section [2.1], which states:“In France current thinking is 2 hours for the fire
and any local applicable time temperature all cases, the minimum requirement is that brigade intervention; after 2 hours it would
curves, with regard to the possible events heavy equipment should not fall down be considered to be unsafe. If the tunnel is
within the tunnel. Hence the early stages of when evacuating users or rescue personnel under a building and in other cases where
the fire development, following the first part are in the tunnel. This means no heavy item protection of property is an important issue,
of the curve, will require a consideration must fall under exposure to temperatures then a longer time may be considered.

Proposed Guidelines by PIARC RWS or the HCinc curve to define design

criteria for different circumstances. The
The proposed guidelines for design criteria Working Group believes that the RWS and
are presented in Table 2.3. This table makes HCinc curves correspond to very similar
a distinction according to the type of traffic levels of fire resistance, and only one of the
(consequently the possible fire load) and two should be used. ISO TC92/SC2 also
the consequences of a structural failure believes that which one you use has no
due to a fire (when the consequences are impact, but considers that the HCinc curve
unacceptable, a protection against a very is a more natural, better choice, should one
severe fire is required – e.g. submerged only be kept. Currently the Working Group
tunnel or in unstable ground; when the proposes that any of these curves can be
consequences are limited, no protection is used, with very similar results.
needed – e.g. tunnel in stable ground).
Table 2.3 uses the ISO curve and either the
Main Structure Secondary Structures4
Immersed or
Tunnel in unstable Tunnel in stable Emergency Emergency exits to
Traffic Type under/inside Cut & Cover Air Ducts5 Shelters6
ground ground exits to open air other tube
superstructure
ISO ISO ISO ISO ISO ISO
Cars/Vans 2 2
60 min 60 min 60 min 30 min 60 min 60 min
RWS/ HCinc RWS/ HCinc ISO ISO RWS/ HCinc RWS/ HCinc
Trucks/Tankers 3 3
120 min1 120 min1 120 min 30 min 120 min 120 min7
Table 2.3 : PIARC Recommendations

Notes :
1
180 min maybe required for very heavy traffic of trucks carrying combustible goods
2
Safety is not a criteria and does not require any fire resistance (other than avoiding progressive collapse). Taking into account other objectives may lead to the following
requirements:
• ISO 60 min in most cases
• No protection at all if structural protection would be too expensive compared to cost and inconvenience of repair works after a fire (e.g. light cover for noise protection)
3
requirements:
• RWS/HCinc 120 min if strong protection is required because of property (e.g. tunnel under a building) or large influence on road network
• ISO 120 min in most cases, when this provides a reasonably cheap protection to limit damage to property
4
Other secondary structures should be defined on a project basis
5
In case of transverse ventilation
6
Shelters should be connected to the open air
7
A longer time may be used if there is a very heavy traffic of trucks carrying combustible goods and the evacuation from the shelters is not possible within 120 min
2.3.2 Pending questions In 2015, revision work on the Eurocodes has Recently, renewed attention has been raised
started, among which the relevant parts for fire in ISO TC92Mid 2017, a decision will be
Introduction of Tunnel Fire Curves into assessment. Recently also through national taken to install a taskgroup, which should
European and International Standards input in CEN TC127, attention was raised to investigate if and how ISO TC92 should
(the Horizontal Group Fire of) CEN/TC250, to start activities on tunnels, and if so, how it
PIARC has contacted the European address the lacking tunnel fire curve in structural should organise these activities in its SC’s.
Committee for Standardisation, in particular Eurocodes. The process to adopt tunnel fire Anticipating on a positive decision as well
CEN/TC250 (”Structural Eurocodes”) and curves has started, but the final implementation as on the taskgroup’s findings, it appears
proposed that a temperature-time curve in the CEN standards, may take a number likely that all 4 SC’s (SC1 “fire initiation and
representative of very severe tunnel fires of years. At the same time, the introduction growth”, SC2 “fire containment”, SC3 “Fire
(either RWS or HCinc) be introduced into the of the supporting calculation rules should threat to people and environment”, and SC4
relevant European standard. be considered for inclusion in the “material” “Fire safety engineering”) are involved in the
dependent Eurocodes, especially EN 1992-1-2 activities. The taskgroup should report and
for concrete. advise by the end of 2017.

Fire Safety Engineering These comments open the field to fire

safety engineering and performance-based classification Type of Vehicle
The important FSE standardisation approach namely new approaches for the
committees and tunnel fire committees, assessment of the safety level. Category 1 Cars only (no HGV)
including CEN TC127/WG8, ISO/TC92/SC2
Category 2 Heavy Lorries (HGV)
and SC4, PIARC D5 and ITA WG6, believe 2.4 ITA Classification
that every tunnel is unique as concerns the Category 3 Petrol Tankers
development of a fire scenario and numerous Based on the information developed in
Figure 2.4 : Road tunnel Categories
parameters are of importance: Guidelines for Structural Fire resistance
• The type and density of traffic, and for Road Tunnels (ITA 2004) it has been
consequently the fire load and its determined that the temperature vs. time The general classification of tunnels as
distribution (area), as well as the possible curves should reflect the typical use of road determined by vehicle type and fire time
fire spread tunnels and a more general classification duration is further expanded by the tunnel
of tunnel fires. Consideration was given type and potential risk of collapse.
• The cross-section configuration, the length
to the types of tunnel structure, cross
and inclination of the tunnel
section, materials and experience of tunnel
• The ventilation design and ventilation operators and designers. The modified
capacity of the tunnel ITA time/temperature plot also classifies
• The possible use of active measures such the fires based upon the use and does
as sprinklers or water mist not consider the fire suppression system
• The roughness of the tunnel surface and or methodology. For ease in design, road
changes in cross-section tunnel categories are presented in the
following Table 2.4 :
• The thermal inertia of the tunnel boundaries,
etc.
Number Tunnel in Tun-nel Exit

Immersed Cut Exit to
category Vehicles unstable in stable Air Ducts to Shelter
Tunnel & Cover Other Tube
Involved ground ground Open
ISO ISO (ii) ISO ISO ISO ISO

1 1-2 (ii)

1 ≥3 (ii) (ii)
2,3 1-2 (iii) (iii)
2 hrs. 2 hrs. 2 hrs. 30 min 2 hrs. 2 hrs.
2,3 ≥3 (iii) (iii)
3 hrs. 3 hrs. 2 hrs. 30 min 2 hrs. 2 hrs.
Table 2 5 : Design criteria for different circumstances

Notes :
(i) : T
he elongation of the design time temperature curve from 2 hours to 3 hours may be considered by tunnel owners who aim to protect the asset, reduce repair costs and limit
the closure time of the tunnel, limiting economical costs of closure and environmental impact of diverted traffic. These longer fire scenarios correspond to a fire with important
propagation to multiple vehicles.
(ii) Personnel Safety is not a criteria and does not require any fire resistance (other than avoiding progressive collapse). Taking into account other objectives (asset protection, limited
costs in case of fire…) may lead to the following requirements:
• No protection at all if structural protection would be too expensive compared to cost and inconvenience of repair (e.g. light cover for noise protection)
(iii) Safety is not a criteria and does not require any fire resistance (other than avoiding progressive collapse). Taking into account other objectives may lead to the following
requirements:
• RWS//HCinc 120 min if strong protection is required because of property (e.g. tunnel under a building) or large influence on road network

2.5 Time-Temperature curve Ventilation speed leading to an increase in temperature.

versus Megawatts Ventilation systems in tunnels are an With a gradient of say 5% the heat and
important part of the holistic fire safety smoke will climb upwards. If in that case
2.5.1 Issues with conversion concept. Full scale fire tests have shown the ventilation goes into the same direction,
that an increased ventilation speed in the the heat will be taken from the fire location
In the fire specifications of tunnel projects, tunnel tube will most likely increase the fire even quicker, reducing the temperature
the performance of the fire protective lining size and can potentially induce fire spread development at the fire location.
and other passive fire protection measures, from one vehicle to the other. The maximum
is described. heat release rate for solid materials can Time/duration of the fire
increase by a factor of 1.4 – 1.7 compared Depending on the amount and type of
Passive fire protection systems are first of with a fuel burning outside a tunnel. The fire combustible materials being involved in
all based on a design fire curve in terms of growth rate may increase much faster, or the fire, the fire duration will be influenced,
temperature development over time as the by factor of 6 or more. By increasing the along with the temperature rise in the first
thermal attack to the system. On the other ventilation speed additional oxygen is fed to minutes.
hand there are the thermal failure criteria the fire source and thereby tilting the flames
of the structure or system that requires so the risk for fire spread between vehicles 2.5.2 Concerns about megawatt output
protection, described as maximum increase. A slower ventilation speed of bus fires
exposure temperatures to certain elements reduces the fire size and fire growth rate but
of the structure. The required thermal the duration of the fire will be prolonged. There is a growing concern about the rather
protection can be selected using these low MW predictions of bus fires which are
parameters. Ventilation speed also influences the gas listed in several national and international
temperature in the tunnel. For a given standards and guidelines. Very little research
It is therefore imperative to prescribe fire size and fire duration increasing the has been undertaken so far to substantiate
the selected design fire curve in the ventilation speed will either increase or these figures whereas bus fires in practice
fire specifications of the tunnel project, decrease the gas temperature. In this case have indicated to have larger MW outputs.
along with the thermal failure criteria. The the overall effect of increasing the ventilation The authors therefore have little information
thermal failure criteria can sometimes be speed may be a lower thermal attack to the to analyse in this document but felt that this
derived from fire testing procedures and system. Therefore it is not clear, whether concern at this stage has to be shared with
standards. mechanical ventilation will increase or the industry.
decrease the thermal attack. On top of that, buses are increasingly
In some cases only the Heat Release propelled using new (more environmental
Rate (HRR), along with the fire duration, Location of the fire in the tunnel friendly) energy carriers, such as CNG. This
is mentioned in the project requirements, comes with additional risks, such as jet fires
If the fire is located near the entrance or
without any guidance as to the time – [2.7].
exit of the tunnel tube, the heat can escape
temperature development. This brings
from the tunnel and can dissipate into the
up the question of how to convert a HRR 2.5.3 Recent research and modelling
surrounding atmosphere. Should the fire be
figure to a design fire curve. For example,
located in the centre of the tunnel, the heat
which fire curve represents 100 MW for 4 If all the above parameters would be
is trapped and will start to heat up the walls
hours? known, an advanced Computational
and ceiling, which in return will radiate the
heat back into the tunnel. Fluid Dynamics (CFD) calculation would
In fact there is no direct physical relation have to be made in order to come to
This is also related to the length of the
between HRR and time-Temperature, an understanding of the temperature
tunnel. In a short tunnel the heat can quickly
so to answer the question, the following development in a certain fire scenario. The
find its way to one of the exits, decreasing
issues need to be addressed: complexity of CFD calculations, especially
the temperature in the tunnel.
when aimed at accurate predictions of
Tunnel Height and Width temperatures, makes them not always the
Gradient
A fire in a high tunnel will build up less heat first choice when doing performance based
The gradient or slope of the tunnel
as opposed to a fire in a tunnel with a small design. Using tables and predetermined
influences the so called chimney-effect.
cross sectional area. In a large tunnel, the time-temperature curves is the most simple
If the tunnel has no gradient the heat and
air volume that needs to be heated is larger and robust method. There may however be
smoke will spread through the tunnel in the
and also the surface area of the walls and situations or projects, where more efforts
same direction as the ventilation direction.
ceilings is larger, and is therefore able to are needed, and then a CFD may be an
In case of a low ventilation speed the heat
absorb more heat. option. This can be the case if one wants
will build up at the location of the fire,

to investigate a specific type of scenario

involving multiple vehicles under a very
critical infrastructure. The interest may be
in analysing the effects on the construction,
not in the point where the fire started,
but further downstream if the fire spread
to other vehicles. Here CFD is a possible
option. There are also other intermediate
solution available, namely engineering
models that correlate the heat release
rate to the ceiling gas temperature. These
engineering models can be used in pure
performance type of design, and it should
be encouraged that designer starts to use
such approaches. An example of such
model is given below.
Recent research has however revealed the

possibility to calculate the maximum ceiling
temperature, based on some of parameters
as outlined under 2.5.1 thus enabling the
designers to convert their conceptual
tunnel design (based on the Heat Release
Rate curve and other data) to a maximum
expected temperature at the ceiling of the
tunnel. The tables presented in par. 2.3
above, are based on the simplified model.
Such models can be both very useful but

also very susceptible to user error. Amongst
other reasons, the user has to understand
to which maximum ceiling temperature the
model is limited.
Very often the fire load in the models,
including the more advanced CFD modelling
is posi-tioned in middle of the tunnel (no
flame impingement on wall), which in almost
all fire scenarios is not typical.
Another subject that the user needs to be

aware of is to which extent fire spread to
other vehicles is accounted for in the model.
The example of the model developed by

Li and Ingason [2.8, 2.9] is presented in
appendix 2.

3 >> Lining material behaviour
3.1 Concrete • In addition to that, within porous lining than concrete with siliceous aggregates.
materials (concrete, and also passive Taking into account also the moisture content
In the following is briefly described the protective materials), mass transport the specific heat capacity is in the magnitude
behaviour of the lining material during heating takes place under influence of the heating of 0.8 - 1.2 kJ/kgK, with a peak at about
from a fire. The behaviour associated with (moisture and above 100°C also water- 100°C dependent on the moisture content.
the heating itself and the associated loss of vapour). For a water-content of 3% the peak value is
strength and stiffness is discussed separately. in the magnitude of 2 kJ/kgK. Alternatively a
The context of the structural behaviour is Accurate assessment would therefore require coefficient indicating the capacity multiplied
dealt with in section 5. advanced computer simulations of the with the density can be indicated. The
physical phenomena involved, and results in coefficient is for concrete in the magnitude
In the normal load cases the thermal rather complex coupled models to capture 2.3 MJ/m3K, (EC value). The 100°C peak is
properties of concrete are of limited chemical (phase) changes, thermal-hygral in the magnitude of 4.5 MJ/m3K for a water-
importance to the design. However in case and thermo-mechanical physics. content of 3%. In the Eurocode there is a list
of a fire the thermal properties are important of temperature depend correlations for heat
for the transfer of heat into the structures, the 3.1.2 Thermal properties conduction, enthalpy and heat of capacity
increase in temperature, which subsequently for different concretes. Both upper and lower
influences the mechanical properties and the The information required for the determination limits are given. It is recommended to use the
load-bearing capacity. of heating of the lining is: values given there, if the exact knowledge for
• Thermal conductivity (depending on the concrete is not known.
3.1.1 Heating aggregates, temperature and humidity)
For calculations of the temperatures in a
• Specific heat /heat capacity (depending on
The heat inside the tunnel will be transferred tunnel it may be necessary to model also the
aggregates and temperature)
to the lining with a rate depending on the thermal properties of the surrounding ground,
tunnel material. This process of heat transfer • Density (depending on concrete mix) in which case the thermal properties of the
involves terms of convection as well as • Convection coefficient (depending on ground will have to be modelled.
radiation. temperature and boundary conditions)
• Emissivity coefficient (depending on surface In principle it may be possible by use of CFD to
Heat transfer by means of convection under and shape) determine the temperature in the tunnel and
fire conditions can reasonably accurately be in the lining at the same time. This will ensure
• Thermal properties for surrounding ground
correlated linearly to the temperature levels of the modelling of the interaction between the
and other thermal boundaries.
the fire gases, whereas heat transfer by means lining and the gas temperature. However, it
of radiation is related to the temperature to • Decomposing reactions (dependent on makes the CFD model even more complex
the power 4. With a view to assess the tunnel material mix and water content) and makes an extreme detailing necessary.
lining behaviour, in most practical tunnel So in most cases it is practical to separate
fire situations, the heat transfer by radiation The properties will depend on the type of (un-couple) the two calculations.
therefore governs. [3.1, 3.2 and 3.3]. concrete (especially the aggregate, either
normal weight (granite, limestone…) or light Note that the heating is in most practical
Knowing the temperature-time curve and weight). It is noted that aerated concrete may cases sufficiently accurately described with
through that the heat transfer to the tunnel have a rather low thermal conductivity, but a thermal model only, implicitly accounting
lining, further assessment of the conduction these types are of little use to normal tunnel for mass transport (hygral, i.e. moisture and
of heat inside the tunnel lining is possible. construction. Also the cement aggregate vapour).
However, given the rather high temperature ratio and the moisture content may influence
levels to be considered, the behaviour of the the (apparent) thermal conductivity. In A final, but important remark is made with
tunnel lining appears specifically complex: addition the thermal conductivity appears to respect to the fact that due to the migration
• The thermal material properties vary decrease with increasing temperatures. In of the heat front through the tunnel linings,
(significantly, and non-linearly) with normal cases the conductivity is simplified by maximum temperatures may be reached at
temperature (such as thermal conductivity, a double linear curve with values between 2 a point in time after the fire has reached its
and the specific heat); and 0.5 W/mK. maximum temperature, and a decay phase
of the fire has started. This lagging effect may
• Also the thermo-mechanical material
The specific heat capacity for the dry concrete extend to as long as 30-60 minutes, and it
properties show a significant non-linearity
will be in the magnitude of 0.5 to 1.5 kJ/kgK suggests that critical structural behaviour
at elevated temperatures (expansion
increasing with temperature and depending may become apparent long after the fire has
coefficient, and strength and stiffness
on the aggregates. Concrete with limestone died out, or has been extinguished.
parameters);
aggregates tends to have higher capacity

3.1.3 Mechanical properties aggregates, the temperature, time and the components of concrete causes internal
load level. The strain from external forces stresses and will introduce (micro) cracks at
Concrete depends on the reduced stiffness and the temperatures as low as 150°C.
strain compatibility.
During the fire, thermal gradients will Reduction of strength and stiffness
develop in the concrete lining, which over The deformation of the cross sections,
a depth of no more than a couple of cm, the segments and the tunnel structure in The concrete is weakened in compressive
may range up to more than 1000 C. Also general shall be investigated in order to strength and in stiffness due to the heating,
large thermal gradients may develop over find the degree of confinement. Often the i.e. thermal decohesion and thermal
the length of the tunnel, with associated deformations are able to redistribute the damage of the concrete. The reduction of
complex effects of (restraint) to thermal loads and relieve confinement stresses. strength and stiffness is normally expressed
expansion. Particularly for a tunnel lining in curves indicating fc respectively E relative
which will be heated from one side the When the thermal expansion is confined to the value at 20°C for temperatures up
temperature and the constitution of the it will produce a mechanical stress. The to 800°C - 1000°C. Values can be found
concrete on the inside will be very different stress is determined by multiplication of in handbooks or codes, e.g. Eurocodes,
from the conditions at the outside. The the confined expansion with the stiffness at national codes and e.g. CEB. The curves
mechanical properties of each part of the the particular point. It should be noted that found in codes are in many cases a
section will be affected according to the the stiffness will be significantly reduced reasonably good estimate for the remaining
depth and the heating of this particular at the points of the highest increase of strength and stiffness. It should be noted
point. The loading of the individual points temperature. The thermal strain is an however, that the curves depend on
of the section is dependent on the thermal expansion of the concrete, however, the the concrete mix: in Eurocode curves
expansion, different creep/ transient total deformation may in some cases be a are indicated for siliceous aggregates
strain phenomena and influenced by compression. respectively calcareous aggregates. In the
the surrounding stress state. Overall the work of Prof. Glerum, a number of curves
boundary conditions of strain compatibility In contrast with the expansion of the for different aggregates are indicated [3.4].
and static equilibrium will govern. aggregates in fire conditions, the cement Furthermore test results reveal a variation of
paste will dehydrate and therefore shrink. strength reductions [3.5].
Expansion and shrinkage The adverse behaviour of these two
Similar to other materials concrete

will expand when heated. In the most
simplified form the expansion is described
by a coefficient giving a linear relationship
between the increase in temperature and
the expansion. However, for the calculations
of concrete structures subjected to fire
it is relevant to define the expansion by
curves giving the relationship between
the strain and the temperature. This curve
will depend on the type of aggregate, the
aggregate cement ratio and the water
cement ratio. For siliceous concrete the
thermal elongation will be about 0.5 % at
400°C and 1.4% over 700°C. Simplified
linear or double linear relationships can be
found from codes or handbooks.
In addition to the thermal expansion also

transient strain, creep strain and strain
from external forces contribute to the
deformation of the concrete. The transient
strain depends on the temperature change
and the level of load whereas the creep Figure 3.1 : Reduction of compressive strength of concrete at elevated temperatures according to
will be dependent on the matrix and Eurocode 2 (EN1992-1.2) [3.2]

Generally in structural design, permissible To go further, we can distinguish 2 cases: the temperature within the concrete
stress in structural members is assumed continues to rise;
to be within the range of 30% 50% of the a/ if no temporary damage of the • To minimise the concrete damage, a
compressive strength of concrete by the concrete structure is allowed maximum temperature of 380°C is defined;
suitable scale of fire with comprehensive it provides a certain safety compared with
risk assessment. This is the case for example when the the limit of 450°C discussed before
requirement is to reopen the tunnel very
• Large shear forces are concentrated near
The temperature at which concrete quickly after a fire with all the guarantees that
the supports (walls), at places where there
spalling may occur will vary between the the structure characteristics remain fully intact.
is no shear reinforcement, which occurs
condition of the temperature which entire Based on experience, the accepted maximum
especially in the older tunnels.
concrete structure receives, the concrete temperature of 380°C(1)is to be applied to
strength, cement water ratio , the quality concrete and 250°C to reinforcement steel for
Steel :
of aggregate and the mechanical loading. the following reasons :
The relationship between temperature and • It is true that the strength of steel is not
strength, showed in Figure 3.1, would not Concrete : reduced at 250°C, but deflections occur.
have any strength reduction up to 200°C. Due to the permanent load on the roof
•B
oth concrete and lining have a high
of the tunnel, the chance is high that the
thermal capacity. They hold the heat and
Figure 3.1 shows the effect of temperature deflections will be permanent.
even after the fire has been extinguished,
on the strength of concrete. The strength
of concrete is still 100 % at a temperature
of approximate 200°C and ends at 750°C.
The effect of temperature on high tensile
steel is worse than on reinforced steel.
Before 1992 only one safety coefficient to
calculate the safety of concrete structures
built in the Netherlands, and this was
1.7. This means that if the strength of the
concrete is reduced to 100% / 1.7 i.e.
59%, the safety of a structure is 1.0, in the
other words it is no longer a safe structure
(very simplified). The graph shows that
the temperature of concrete can rise to
approximately 450°C before a structure
might fail.
According to Figure 3.1, it can be thought

that hair crack or spalling already exists over Figure 3.2 : An arbitrary example of a temperature development throughout the structure of concrete with fire
the concrete surface, which has already protection material
been affected by high temperature, in the
area of temperature where initial strength
gets lower (decreases). Therefore, if the
requirement is to avoid any resistance loss, Figure 3.2 illustrates a concrete tunnel liner Note that the below concrete related
it is necessary to limit the temperature that with 25 mm of fire protection material on parameters do have a significant influence
the concrete is subjected to a range within exposed surface, cover on reinforcing steel on the temperature development in a
350°C 400°C. In any of these cases, if is 25 mm and slab thickness is assumed concrete structure:
the temperature concrete is subjected to to be 250 mm. Fire exposure is 120 • Type of aggregate being used (calcareous
in a fire is between 1,000ºC and 1,350ºC minutes Rijkswaterstaat (RWS). This graph or siliceous)
(RWS standards), set at planning or is indicative only and cannot be used for
• The density of the concrete
design stages, it is required to take into specific projects.
• The moisture content of the concrete
account the resistance reduction and
eventually provide materials that shield
an approximate maximum temperature of These values are related to the retained strength and durability of (ordinary cast in-situ) concrete. Espe-cially for
(1)
1,350ºC (see below). higher grade concretes, lower values (180-220 °C) may apply.

b/ if temporary damage of the concrete Some of key observations concerning The question is then what can be done to
structure is allowed spalling are listed below : take into account the extent of spalling
• Dense, impermeable concrete is more in the estimation of fire resistance. Some
Concrete tunnel structures may be fire susceptible than low quality concrete. ideas are given:
resistant without the use of fire passive 1. The spalling characteristics can be
•H
ence, the concrete mix and production
protection. In this case, the effects of determined experimentally. This estimation
conditions are important parameters.
high temperature on the materials and has the disadvantage that it is performed
the structures (resistance and stiffness •S
palling can occur both for high strength
in a relatively advanced stage of the
reductions, thermal dilatation…) have concrete and classical concrete.
construction. Due to the size effect, and the
to be carefully taken into account in the • The diffusion of water and the water production conditions, a full size sample
design. content are important. should be taken from the production and
•T
he aggregates influence the spalling the sample should have a reasonable age,
It is important to keep in mind that even (calcareous aggregates are reported as the spalling is dependent on the water
unprotected, a tunnel may be able to to give more spalling than siliceous content. A new development is mobile
guarantee the required level of safety aggregates). furnace which can be installed in an existing
during a fire. The main drawback is that tunnel or at a yard nearby the construction
the concrete will be damaged by heating •T
he arrangement of reinforcement
influences the spalling especially in location of the tunnel, for a segmented
and require more extensive repair works. tunnel for example. The mobile furnace
In such a case, the tunnel owner should be normal strength concrete but it does
not prevent the spalling (contrary to aims to define the spalling sensitivity of the
aware that, after a fire, the tunnel may be concrete, providing interface temperature
closed for a longer period. Especially for indication in some codes).
limits at time intervals during the fire
immersed tunnels, fire induced cracking • Spalling of unprotected concrete exposed
exposure period.
at the unexposed side might occur, to rapid increase of temperature (RWS,
HCinc) commences within minutes. 2. The spalling can be assumed. It has
which initially may be acceptable, but
earlier been indicated in codes etc. that
due to durability issues over time lead to • Fires with rapid increase of temperature
spalling could (in case it is taken into
unacceptable situations. (RWS, HCinc) are worse than fires with
account) be assumed to be limited to the
gradual increase of temperature.
cover, i.e. 30 - 50 mm. This has in case of
•M
echanical pressure (compression) on especially the Channel Tunnel shown to be
3.1.4 Spalling the structure is reported to make the too little. However, in place of other design
spalling worse. basis one can assume a certain spalling
Whereas the reduction of strength and
•T
he geometrical shape of the element e.g. 50, 100 or 150 mm and test the
stiffness is relatively well described,
influences the spalling pattern. structure towards this spalling damage.
the spalling is an area for which the
•T
he occurrence and course of the The spalling can in turn be confirmed or
mechanisms are still discussed. A practical
spalling is a random process if studying updated by experimental testing.
engineering model for determining the risk
of spalling and the possible thickness of the individual fragments. 3. It can be aimed to prevent or limit
spalling does not exist yet. •S
elf Compacting Concrete is reported to the spalling. A classical idea for existing
suffer more from spalling compared to structures is fire protection, in form of
On the other hand the topic of spalling is regular concrete. boards or sprayed-on material. Especially
important to take into account. There is not pre-mounted material has disadvantages
•T
he thermal gradient that the interface in connection with segmental lining, as
much point in discussing the reduction of
will be exposed to impacts on the it may be damaged during erection or
the strength and stiffness if it is not known
spalling behaviour. will hinder the erection rate. In any case
whether 50, 100 or 150 mm or more may
spall during the fire. •C
oncrete structures which are also fire protected concrete may under
protected against fire, but using circumstances spall, so the fire protection
There are two basic mechanisms, which insufficient insulating capacity (material must have a suitable thickness (see [3.6]).
internationally are considered now as type improper or material thickness For new tunnels, other fire protection
equally relevant to consider, but depending insufficient), can suffer from more measures like admixture of polypropylene
on specific cases one mechanism may (explosive) spalling. and steel fibres can be applied. These
govern the spalling phenomena, or measures are not discussed in this
initiate it: the water vapour pressure clog A realistic practical engineering estimation document. Reference is made to relevant
consideration and the (restraint to) thermal of the spalling is not common practice and research [3.7].
expansion. may not be possible with analytical tools.

3.2 Reinforcement Mechanical the reinforcement is reduced at a higher rate

as illustrated in figure 3.5.
Thermal Similar to the concrete the strength and
stiffness of the reinforcement is reduced due Simplifications of the strength reduction
The thermal behaviour of the reinforcement to the heating. For many types of structures curves are indicated in different handbooks
can be described by the conductivity, the the load bearing capacity during a fire is and codes (the curves of figures 3.3 to 3.6
heat capacity, etc. similar to the parameters directly dependent on the moment capacity are taken from Eurocode 2 Part 1-2). These
indicated for the concrete. and thereby on the remaining capacity of the indications will in most cases be sufficient; if
reinforcement. For tunnels the capacity is more information is necessary the producer
The detailed description of the heating often determined by a combination of bending of the reinforcement may have test results.
and heat flow in the reinforcement can be and normal force (see section 5). Otherwise particular tests will have to be
important for the analysis of details around made.
the reinforcement or for heavily reinforced The reduction of the strength and stiffness
parts of the tunnel. is dependent on the type of reinforcement: The reinforcement will to some extend regain
cold deformed reinforcement will lose more its strength after cooling. The hot rolled
For less reinforced parts of the tunnel, the strength at high temperatures than hot rolled reinforcement will regain most of its yield
calculation of the heating of the reinforced reinforcement. strength and ultimate strength. The cold
concrete may often be simplified as a Examples of the yield strength reduction are deformed reinforcement will regain most of its
homogeneous material. indicated in figures 3.3 and 3.4. The stiffness strength up to a heating of 300°C, for heating
and the stress-strain relationship of the up to 600°C about 60% of the yield strength
reinforcement in general will be affected by the and ultimate strength will be remaining.
heating. As for the concrete, the stiffness of
Figure 3.3 : Relative strength of class N reinforcing steel at elevated temperatures Figure 3.4 : Relative strength of prestressing steel at elevated temperatures according
according to Eurocode 2 [3.2] to Eurocode 2 [3.2]
Figure 3.5 : R
elative modulus of elasticity of reinforcing steel at elevated temperatures Figure 3.6 : Relative modulus of elasticity of prestressing steel at elevated
according to Eurocode 2 [3.2] temperatures according to Eurocode 2 [3.2]

When subjected to very high temperatures and certified products which can withstand and dense smoke under fire conditions,
the steel may melt or decompose. design fire curves which can withstand fires which make them unsuitable for use in
Details should be particularly observed: that range between ISO and RWS. For each tunnels where life safety is an important
welds will also lose strength and corrosion tunnel project the suitability of the use of the consideration.
protection by paint or similar may influence intended product and system needs to be
the behaviour of the structure during the checked against the project requirements. 3.3.2.2 Magnesium Oxy-Chloride
fire. based materials
The common products are:
Thermal expansion • Calcium Silicate boards In the general building industry, the use
of Magnesium Oxy-Chloride based board
• Calcium Silicate Aluminate boards
The thermal expansion of the reinforcement materials is rather common. Several of
is in most cases modelled by a linear • Vermiculite / cement spray on systems the manufacturers of fire protective board
or double linear temperature- strain • Perlite / cement spray on systems materials in the market, either produce such
relationship. The expansion coefficients • Composite panels boards or have access to this technology.
are moderately influenced by the type of Magnesium Oxy-Chloride materials are also
reinforcement. The coefficient is higher for Reference is made to special studies and known as “Fibre reinforced magnesium”
pre-stressing cables than for reinforcement product information in section 5 [3.4, 3.8]. and “Magnesium oxide, silicates and other
bars. The coefficient for reinforcement is in Fire protection materials are discussed in additives” which all contain Chloride.
the same magnitude as the initial thermal sections 5.12 and 5.13.
expansion coefficient for the concrete. However, for technical and life safety
By means of passive fire protection the reasons mentioned below, it has to be
Bond of reinforcement heating can be reduced. The types of fire noted that Magnesium Oxy-Chloride
protections vary from organic to inorganic boards are not suitable for application in a
Due to the heating, the reinforcement materials, sprayed-on material and boards. tunnel environment.
will gradually lose its bond. The reason is
mainly cracking in the concrete and the 3.3.2 Non-suitable fire protection for a/ Corrosive effects of leaching
reduction of the strength of the surrounding use in tunnels Chloride ions
concrete. Curves indicating the relationship Magnesium Oxy-Chloride present within
between the bond and the temperature are 3.3.2.1 Paint products and systems the board matrix will have a detrimental
given in the literature, but the exact models corrosive effect on the anchors, steel
are still disputed. It has been observed that Fire isolating paints also exist but these components in the tunnel and steel
ribbed steel remains more of its bond up to are not suitable for use in tunnels. These reinforcement in the concrete, especially
temperatures of 400°C - 500°C compared Inorganic Coatings, also known as in high humidity environments negatively
to smooth steel. At temperatures over Intumescent Coatings can be either Solvent impacting on the long term durability of the
600°C, no or very little bond will be left. based or Water based and rely on the ability completed tunnel facility. This is associated
of the product to react under fire conditions with the leaching of Chloride ions when
If possible structures should be designed and swell to create an insulating char. the Magnesium Oxy-Chloride material is
so that the reinforcement is bonded at the These products would be unable to react exposed to moisture, high humidity levels or
side of the structure that is less subjected in a timely manner to keep pace with the water. Reinforcement steel, steel brackets,
to fire. rapid temperature development associated anchors, jet fans, cable trays etc. will be
with tunnel fires. Therefore the use of such exposed to Chloride concentrations to the
For the design and analyses of details coatings is not recommended. extent that corrosion and/or premature
such as joints and fixations a closer study failure is inevitable.
of the bond of the reinforcement might be In addition:
relevant. For larger structural elements it Reports have been available in the public
I) Solvent based products are combustible
may not be necessary to study the bond domain in this regard for years and most
and health and safety hazard during
in detail. recently a further critical report has emerged
application needs to be considered
from an expert consultant in Australia on
3.3 Protective measures II)
Water based products most typically
this very topic [3.9, 3.10, 3.11].
cannot cope with heavily laden moisture
3.3.1 Commonly used product types conditions often present within a tunnel
Such detrimental effects on the durability
requiring the additional application of a
of the tunnel are a concern with owners,
The industry of passive fire protection water protective topcoat (Solvent Based)
engineers and contractors.
materials offers a range of suitably tested They are known to produce toxic fumes

b/ Impact of Hydrochloric acid (HCl) on it is exposed to high humidity due to the

tenable environment dissolution of the salts it consists of. If metal
During fire conditions, Chloride is being parts, such as galvanized steel, fasteners
released from the board, exposing tunnel and flashings are in contact with MgO-
users, who are trying to escape from the fire boards, they will start to corrode within
scene, and emergency response teams, to short time, leading to safety problems in
Hydrochloric acid (HCl). the structure. Sheathing with MgO boards
Such detrimental effects on the tenable are now being replaced by other types of
environment and life safety of tunnel users materials in a great number of buildings in
cause serious concerns with owners, Denmark.”
engineers, and emergency response teams.
Humidity levels in tunnels are often rather
c/ Independent research conclusions high, due to, for example:
The Technical University of Denmark and • Water leaks of the tunnel structure
Bunch Building Physics ApS from Denmark • In case of precipitation, moisture will be
have issued a paper which is attached in carried into the tunnel by vehicles
annex [3.9]. Their recent (August 2016) •W
ashing activities to clean the tunnel
research concludes as follows: linings
• Meteorological circumstances.
“The experiments described in the paper
seem to indicate a relative humidity level of
84% RH as a limit above which the MgO-
boards that have been tested begin to
absorb excessive amounts of moisture from
the surrounding air. Since this and higher
values of relative humidity are typical in
locations where exterior sheathing is to be
used, it can be stated that MgO-boards
cannot be a suitable product for this use.
This is supported by the numerous examples
of failures, which have been seen in recent
years, where such boards leak salty water
that cause damage on adjacent construction
members of wood or metal, and which even
over time lead to disintegration of the MgO-
board itself.”
They continue to conclude :
“For this reason, it can be stated that MgO-

boards are not suited as sheathing in exterior
facades or any other application where the
boards are in contact with a moist climate. At
RH above approximately 84% RH the MgO-
boards will form drops of salty water on the
surfaces, and this water will be absorbed
in wooden structures in connection with
boards and water absorption will increase
and can lead to mould growth on wood.
The MgO-board itself is also sensitive to
mould growth due to the content of organic
material. Furthermore, the MgO board will
itself also be disintegrated over time when

4 >> Tunnel classification
4.1 General 4.2 Tunnel Classification Type 4B Immersed Tube; Supply air above
Description the roadway, no separate exhaust
Tunnels are classified in regard to the original Type 4C Immersed Tube; Supply air below the
design the type of construction and the The tunnel description is intended to roadway, no separate exhaust
physical properties of the tunnel in regard to simply classify tunnels into numerous Type 4D Immersed Tube; Jet fans above the
the method of ventilation, egress and potential classifications that will allow the tunnel type roadway, no separate exhaust or
for collapse as a result of a fire event. Tunnels to be placed in a database for rapid retrieval supply Type
are generally classified as follows: and to standardise the de¬scription of Type 4E Immersed Tube; Immersed tube
• External configuration (circular, box section, tunnels. The most common types of tunnels tunnel with cross passages
arch) are as follows: Type 5 Tunnel with external egress tunnels
• Structural lining type Type 1 Circular; Exhaust above; supply (any type of construction)
• Method of construction air below roadway Type 6 Urban tunnel with structure above
• Type of environment for tunnel construction Type 1A Circular; Supply & exhaust below tunnel (any type of construction)
(stable, unstable ground) the roadway
• Ventilation configuration Type 1B Circular; Supply over the roadway, 4.3 Tunnel Classification for Risk
• Location (under/near other structures, in no exhaust Analysis
water course) Type 1C Circular; Exhaust over the
• Types of emergency access (rescue tunnels, roadway, no supply The classification of tunnels as shown here
cross passages, emergency shafts) Type 1D Circular; Jet fans over roadway, no are intended to allow the owner/operator
External configuration of tunnels is divided up separate supply or exhaust to easily evaluate the degree of sensitivity
into the following subsets: Type 1E Circular; Supply and exhaust over towards risk associated with a fire in the type
• Circular the roadway of tunnel structure that one is evaluating. The
• Box (Square or rectangle) Type 2 Arch; Exhaust over the roadway key elements for protection are those that are
• Arch Type 2A Arch; Supply air over the roadway to be protected from collapse or are necessary
• Twin Tube (Immersed Tube Tunnel), ITT) Type 2B Arch; Jet fans in arch, no separate for the safe evacuation of the public and for
supply or exhaust protection of fire-fighters. Tunnels that are
The structural lining type is divided up into the
Type 2C Arch; Supply air below the located in urban environments and/or are in soft
following subsets:
roadway unstable ground have the greatest potential for
• Cast-in-place concrete
Type 2D Arch; Jet fans in arch, supply air risk associated with collapse or the potential
• Shotcrete
below the roadway to effect other nearby structures (types 4,5,6),
• Precast concrete segmental
Type 2E Arch; Supply and exhaust below therefore more attention must be made to the
• Precast concrete
the roadway protection of the structural elements from fire.
• Exposed cast iron/ steel segmental
Type 2F Arch; Supply and exhaust above
• Concrete covered cast iron/steel segmental
the roadway The information as to classification of tunnels is
• Dimension Stone (Asher stone)
Type 3 Box; Exhaust above the roadway, for informational purposes only and each tunnel
• Masonry (brick)
supply air under type must be classified according to local codes
Type 3A Box; Exhaust over the roadway, and ordinances. The intent of this classification
The Medium or environment that the tunnel is
no separate supply air system is to standardise classifications to
constructed is divided up as follows:
Type 3B Box; Supply air over the roadway, allow local codes and ordinance to adequately
• Stable ground
no separate exhaust evaluate the potential risk and thereby develop
• Unstable ground
Type 3C Box; Supply air below the appropriate guidelines for protection of the
• Bedrock
roadway, no separate exhaust structural elements as described in Sections 5
Type 3D Box; Jet Fans in roof, no separate and 6 of this document.
The type of ventilation means of egress and
supply or exhaust
proximity to other structures is a variable for
Type 3E Box; Supply and exhaust below
each type of tunnel. These variables have been 4.4 Tunnel Classification
the roadway
classified to illustrate each type of ventilation Sketches
Type 3F Box; Supply and Exhaust above
configuration and are best illustrated in the
the roadway
attached figures. The use of external rescue The following sketches are for informational
Type 4 Immersed Tube; Exhaust above
tunnels, emergency access shafts and cross purposes only and are representative of the
the roadway, supply air under
passages and proximity to other structures classifications illustrated here. Variations in
Type 4A Immersed Tube; Exhaust above
are modifiers to each general type of tunnel these configurations are common and are to
the roadway, no separate supply
illustrated. be documented as discussed in Section 4.2
air

Structural Fire Protection for Road Tunnels




The following table is provided to illustrate typical elements of a tunnel to be preferentially protected from rapid heat rise that would cause
collapse or an unsafe condition in the tunnel. The protection should be based on the criteria provided in this document and after consultation
with local Authorities of applicable codes.
Tunnel Lining Unstable Lining Stable Roadway Duct Cross Fan Ceiling
Ceiling
Type Ground Ground1 Slab Walls Passages2 Anchorages supports
1 X X X X X
1A X X X X
1B X X X X
1C X X X X
1D X X X
1E X X X X X
2 X X X X
2A X X X X
2B X X X
2C X X X
2D X X X X
2E X X X X
2F X X X X X
3 X X X X X
3A X X X X
3B X X X X
3C X X X
3D X X X
3E X X X X X
3F X X X X
4 X X X X X
4A X X X X
4B X X X X
4C X X X
4D X X X
4E X X
53 X X X X X X X
64 X X X X X
Table 4.1 : Typical levels of protection based on tunnel type

Notes : 1
. SG = Stable Ground and is assumed to be constructed in stable rock
2. Cross Passages used only for twin tunnel construction
3. Interior protection dependent on ventilation system refer to Type 3
4. Interior protection dependent on type of ventilation refer to Type 1

5 >> Structural elements
5.1 General classed as hazardous for the purposes of Level N0

transport).
The following discussion is focused on No risk of progressive collapse in the
the types of linings used for road tunnels Such fires are characterised by modified event of a local failure. The loss of one
and may be applicable to other types of hydrocarbon time/temperature curves element should not result in a transfer of
tunnels and underground structures. This such as the Rijkswaterstaat (RWS) and load, which is likely to cause other parts of
discussion is in general and is an attempt Hydrocarbon Increased (HC inc) curves the structure to fail.
to illustrate the various behaviour of lining described in 2.3.1, which reach 1200°C
subjected to intense heat from a single fire within 10 minutes and may reach 1300°C Level N1
event. or 1,350º shortly afterwards.
This level corresponds to the majority of
5.2 Structural Elements Materials used to provide structural fires excluding the most violent ones. It
fire protection in buildings may not be should be used for structures which are
Construction materials used for main suitable for use in tunnels because they important for emergency action to take
and secondary structures must be non- may fail (by either melting, burning or place.
combustible or very little combustible disintegrating) at temperatures above
(Euroclass A1 or Euroclass A2, s1, d0) 1,200ºC. Regardless of the applicable In tunnels where all types of vehicles
from the point of view of reaction to fire. design time-temperature curve in a are allowed, level N1 corresponds to
Lightweight roof structures may have tunnel project, the melting temperature of resistance to the ISO curve during 2 hours.
less severe fire reaction requirements (i.e. structural fire protection materials shall be
Euroclass C) since the loss of these does at last 1350°C. Level N2
not represent any risk to safety provided
that fire propagation risks are limited 5.3 Objectives and Fire This level corresponds to high intensity
(class A2 when tested to EN13501). All Resistance Levels fires, with a very quick development. It
materials used for internal fire proofing should be required for structures which
linings must be non-combustible (class Fire resistance required for structures must resist the most violent fires during
A1) when tested to EN13501. and equipment is designed to provide the the period required to allow evacuation
following main objectives: and action by the emergency services.
This section specifies the minimum level • P rotection of users inside the tunnel for
of fire resistance of materials to ensure the time taken for them to reach the In tunnels where all types of vehicles
the safety of persons, including the exit. are allowed, level N2 corresponds to
emergency and rescue services engaged resistance to the HCinc curve during 2
• Maintain a safe environment for the
in their work. The tunnel owner may hours.
emergency and rescue services.
specify higher levels of resistance in order
to provide improved fire resistance for the • Maintain electricity supply and Level N3
tunnel and limit both the extent and cost communications on either side of the
of repairs and closure period following a fire. This level corresponds to the most onerous
fire. • P revention of flooding by surrounding fire exposure conditions and applies to
ground water into the tunnel or structures, which must resist the most
Time/Temperature curves are to be used catastrophic collapse of the structure. violent fire throughout the prescribed
as appropriate justification of the fire • P rotection of structures or buildings, exposure period.
resistance of structures and certain items which are in close proximity to the
of equipment. tunnel. In tunnels where all types of vehicles
are allowed, level N3 corresponds to
• Maintain ventilation capacity and design
Fires in which the temperature rise is resistance to both the ISO curve during 4
ventilation requirements.
relatively slow, but which may have hours and the HCinc curve during 2 hours.
long durations are characterised by the
5.4 Fire Resistance Levels
standard ‘cellulosic’ Time/Temperature Tunnels Reserved for Passenger Cars
curve as defined in ISO 834.
In order to meet these objectives without
Where only passenger cars and vans are
unnecessary additional costs, the
Fires involving heavy goods vehicles allowed, with average fire loads of 7MW
following four levels of fire resistance from
may have a much faster temperature each, levels N1, N2 and N3 are identical
the French Inter-ministry circular 2000-63
rise, particularly if they involve highly and correspond to resistance to the ISO
[5.1] may be considered:
combustible materials (even if they are not curve during 1 hour.

5.5 Evidence of Fire Resistance slab; the structural stability of an adjacent tube
of protected structures • The fixation method of the tested fire when there is direct communication or
protection material (type and number escape to the outside of the tunnel
The level of fire resistance of structures of anchors, wire mesh, etc.) which shall This level shall also apply if local structural
and equipment must be proven by be identical to the one installed in the failure is likely to cut a ventilation duct
testing, reference to previous testing, by tunnel; which is important to allow action by the
calculation or by a combination of all three emergency services.
• The presence of joints in between at
of the above.
least two panels when board protection;
Level N2
during the fire test, any thermal leaks
Extrapolation based on specific testing
through these joints will be assessed.
should not be done, and also special This level applies when the structure is
care should be taken when interpolating • In case of spray materials, the number required to maintain the stability of another
between different tests representative for of applications (amount of layers) tube or separation from it when there is no
other fire scenarios (e.g. addressing a when preparing the test specimen; this direct communication with the exterior.
cooling down phase adequately). amount of layers must be respected
while applying the spray material in a real This level shall also apply if local structural
In Europe, the European norm EN 13381-3 tunnel.; failure is likely to cut a ventilation duct
may be used to determine the contribution • The temperature recordings by which would compromise safe shelters
of fire protection to the fire resistance of thermocouples; they shall be done at least: and their access-ways to the exterior.
structural concrete members exposed - At the interface in between the concrete
to the ISO curve. This EN standard is and the fire protection material; Level N3
recommended also outside Europe, but - At the bottom of the reinforcement;
locally other standards might apply. - On the non-exposed face of the This level applies to immersed tube tunnels
concrete slab. or tunnels located below the water table
Concerning higher temperature-time which are at risk of flooding in the event of
fire curves such as RWS and HCinc, no The above two documents are aligned local structural collapse. This also applies
standard exist. However several guidelines on quite a few topics, but also there if local failure of the structure is likely to
have been published to provide fire test are apparent differences, e.g. the result in catastrophic collapse, damage at
method for RWS and HCinc. References dimensions of the slabs, the evaluation the surface or likely to affect the stability of
are made to: of the thermocouples readings (average an adjacent structure.
he
•T fire testing procedure and maximum temperatures), furnace
2008-Efectis-R0695 (September 2008) pressure and moisture conditions. More The Dutch Ministry of Transport
[5.2] which is the first published RWS harmonisation on international level is (Rijkswaterstaat RWS) has minimum
test method, and adopted in NFPA 502 highly welcomed. requirements for fire safety measures
[5.3]; in tunnels, amongst which thermal
5.6 Minimum Requirements for protection of the structure is a mandatory
•T
he CETU guidelines on passive fire
Main Structures requirement.
protection systems (published in 2013
in French and updated and translated in
The main structure of the tunnel must 5.7 Lining, Material Types and
English in 2017) which provide another
satisfy Level N0 when local failure Fire Behaviour
HCinc test method for applied protection
alongside the source of the fire will have
to concrete members [5.4]. These
no harmful consequences for the safety The following types of (concrete) lining can
guidelines have been written by CETU
of users or cause structural failure. In be considered:
and the 3 French approved laboratories
other circumstances, a higher level of fire •H
orseshoe Tunnels
(CERIB, CSTB, Efectis France). The
resistance must be provided. - Cast in-situ concrete lining
proposed test method is based on the
ISO test of the EN 13381-3. •C
ircular Tunnels:
Level N1
- Cast in-situ concrete lining
These publications highlighted the - Segmental lining with gaskets and
This level applies to structures supporting
importance of: packers.
a roadway or an area, which is accessible
•T
he dimension of the tested concrete - Cast iron linings with lead seals
to the public above it.
slabs used for the application of fire •B
ox/rectangular Tunnels
protection materials; It also applies when stability of the - Cut and cover tunnels
•T
he type of concrete used for the tested structure is necessary in order to maintain - Immersed tunnels including joints

Concerning damage to the lining of a to the magnitude of the load at a specific the stiffness is reduced most. Therefore
circular (and horseshoe) tunnel, the load is point of time. The fire will influence the load the thermally induced load may have its
mainly transmitted as compression in the due to the thermal expansion and the load largest contribution from the middle range
hoop direction. The parameters affecting may influence the spalling and thereby the temperature increases.
the remaining strength of the tunnel are: remaining strength.
• amount of segment remaining after The stresses from thermal expansion come
spalling, which is impossible to predict or Water and Ground Load only from restrained expansion; the elastic,
determine accurately plastic and stiff body deformations tend to
The water and ground loads are generally relieve the thermal loads on the structure.
• temperatures reached during the fire and
regarded as unaffected by the fire. However, Particularly segmental lining has the ability
consequential concrete strength loss
for very long duration fires the ground may to deform. Interaction with the stiffness
• tunnel deformations and tunnel - ground be heated so much that the load is affected of the ground is particularly important for
interaction (for bored tunnels). Strong heating may segmental lining in relatively soft ground.
• possibility of rehydration of cement also influence the water pressure. For most The weakened stiffness of the ring will
• position of damage within the ring relative fires and especially for thick concrete lining transfer more load to the ground and
to possible disturbed ground outside the the heating at the outside of the tunnel deformations will be allowed to relieve the
lining (extrados) will be marginal. thermal stresses.
Also here it is worth to notice the boundary
• squatting of the ring.
Water and ground load can be taken from condition of strain compatibility.
design assumptions, from geometrical
For a rectangular tunnel, where the load
information or from measurements. The In the concrete cross section the areas with
is transmitted mainly by bending, the
load will vary along the alignment. As the the highest stiffness will attract the loads,
parameters affecting the strength of the
design of the segments will be uniform or which is fortunate, as the strength is also
tunnel are:
stepwise uniform the fire resistance will available here.
• the possibility of spalling revealing the depend on the location in the tunnel.
reinforcement Moment Action
• the temperature reached in the Water- and ground load will result in normal
reinforcement during the fire (with or force and moment in the lining. For circular Ideally bored tunnels are subjected to hoop
without spalling) and horseshoe shapes tunnels it is aimed load, i.e. a unidirectional load (the term axial
• the amount of concrete remaining to that the load is transmitted as normal load is used even though a ring is not an
transmit normal force force as far as possible. For rectangular axis). But due to limited (small) overburden,
tunnels the moment action will be more variation in ground stiffness, cross passage
• temperatures reached during the fire pronounced. joints etc. the segments are subjected to
and consequential loss of strength and a certain moment. These moments are
stiffness. Thermal Load generally not a problem for the relatively
• ability to redistribute the load after damage thick segments. However, during the fire
to part of the tunnel (creation of hinges) A thermal load derives from the restricted the cross section is decreased, the strength
• possibility of rehydration of cement expansion of the material due to the heating. in the remaining cross section is decreased
and the moment capacity is therefore
5.7.1 Loads Distinction must be made between 1. decreased. More important, due to the
Internal stresses, which may be contributing spalling and the heating at the intrados,
In order to discuss the behaviour of the fire to the explanations of reduction of strength the line (curve) of action of the hoop force
damaged structural element it is necessary and in some theories to the spalling and 2. moves towards the extrados. The contact
to deal with the loads on the structure. On The external overall loads on the segment, point between the segments is at the
a tunnel the main loads are normally water which are added to the water and ground centre line. This means that the segment is
load and ground load. For a discussion of load. subjected to an increased moment.
fire resistance also the thermally induced
load must be taken into account. In addition to the deformation from thermal
expansion also transient deformation, creep In turn the moment will cause a deformation
It shall be noticed that the definition of fire and deformation due to loading and loading of the segment (in interaction with the
resistance is duration (time) of a specified fire, history should be taken into account. stiffness of the ground) and the contact point
which results in a reduction of the strength It is particular important to take into account will move slightly towards the extrados, but
of the structure to a level corresponding the reduced stiffness of the concrete. At the this leads only to a slight reduction of the
position where the temperature is highest moment.

For bored tunnels ground-structure load after damage at mid-span, whereas for combinations of normal force and
interaction may to some degree relieve wide tunnels e.g. road tunnels may be too steel can be used. The calculation is due
some of the loads occurring during a wide to carry the redistributed load. to the heating complicated by the fact
fire. For immersed tunnels the tunnel- that the segment does not have the same
ground interaction is normally in terms In practice it is common to fire protect properties all over the cross section.
of the (hydrostatic) ground load and the the roof and the upper part of the walls The capacity can be expressed as the
foundation. Cut and cover tunnels may have for immersed tunnels by fire isolation maximum uniformly distributed load or
similarities with bored tunnels or immersed material. The fire protection prevents both simply by a relative figure, which is 100%
tunnels depending on their shape. unacceptable heating of the reinforcement at the start of the fire.
and also spalling. When using performance-
For rectangular tunnels the moment action based approach in structural design or The curve describing the development
is more pronounced, depending on the load fire safety engineering, it is possible to in reduction of capacity will be more
and the geometry of the tunnel. The load determine the surfaces to be protected. pronounced than the one only taking
on immersed tunnels will in most cases be For instance, some tunnels have been normal force into consideration.
dominated by a hydrostatic water pressure, protected only on the ceiling and the upper
which will tend to give compression in walls, part (1 meter to 1,5 meter) of the walls. Biaxial or triaxial loading condition
roof and bottom and negative moments in
the corners of the rectangle and positive Immersed tunnels can in some cased be For the design of tunnel lining it will in
moment at mid-span of walls, roof and pre-stressed by cables in the axis direction. most cases be sufficient to consider a
bottom. Damage of these cables can also result in uniaxial loading. For circular tunnels the
lost functionality of the tunnel. main load will be in the hoop direction
With respect to fire the critical point is often and for rectangular tunnels the main
the centre span of the roof, as it often will 5.7.2 Segmental capacity load is normal force perpendicular to
have a positive moment activating the inside the tunnel axis and moment around the
reinforcement in tension and at the same Normal Force axis. The normal forces in the tunnel axis
time this point of the tunnel is subjected to direction come from forces built in during
the maximum heating of the fire. In the ideal situation the load on a bored construction and prestressing.
tunnel segment is a perfect hoop load,
The fire will heat up the concrete and and the product of the area and the The capacity of a bi-axially loaded
reinforcement and the moment capacity factored strength determines the capacity concrete will be higher than a uniaxially
will gradually decrease as the strength and of the segment. In case of a fire damaged loaded structure, so the consideration
stiffness of the inside reinforcement is lost. cross section the spalled area is of of uniaxial conditions will be on the safe
course disregarded and the heated areas side.
In case of spalling the reinforcement may be contribute with their remaining strength
directly exposed to the fire and the moment as far as the strain compatibility allows it. For consideration of details the biaxial
capacity is immediately lost. As the spalling The capacity towards normal force can be and even the triaxial conditions might be
may take place within the first minutes of expressed in terms of MN or in a normative worth to take into account. An example is
the fire, this type of structure is vulnerable value, which is 100% at the start of the fire. the joints of the segmental lining.
towards fire damage.
A curve describing the development of Shear and splitting
Even though the moment capacity at mid the remaining strength during and after
-span is reduced to nearly nothing, it may the fire can be established. The curve At the joints the segments have shear and
not lead to total collapse of the tunnel. will have a steep decrease at the time of splitting forces. The segment is normally
The mid -span may be regarded as hinge spalling, a more gradual decrease during reinforced in this area for these forces.
and the loads may be carried by increased the remaining heating and the decrease is
negative moment at the corners. expected to continue some time after the The calculation of the shear and splitting
fire has ceased. is following the same principles as
For immersed tunnels further development design of cold joints. However, with the
of the damage will result in flooding. For cut Normal Force and Moment reduced strength of the concrete and
and cover tunnels the consequences will reinforcement this detail comes out as a
depend on the use of the surface over the Due to the moment action mentioned critical point in the calculations.
tunnel. in the section concerning the load, the
Narrow tunnels e.g. single track railway capacity will be further reduced. The well-
tunnels may have the ability to carry the established calculation of concrete or steel

Some complications occur in the calculation It should be noticed that the structure in 5.8 Linings
of a shear failure line through strongly varying some cases may have redundancies which
strengths and stiffness and from the fact that can ensure fire resistance even after failure 5.8.1 Concrete Linings
the shear/splitting reinforcement is heated of a single element. For rectangular tunnels
so much at the intrados side that the bond reference is made to the description above. Concrete linings generally fall into two
is doubtful. categories, cast concrete and pre-cast high
It must be noticed, that a number of the strength concrete segments or beams. Each
The capacity of the joint observed in real fire parameters used in the calculation of the type will behave differently under fire exposure
benefits from the fact that the spalling tend fire resistance have large uncertainties. It conditions.
to be substantially less (or not occurring at is advisable to evaluate the importance
all) near the joints. of the uncertainties. This evaluation can Cast concrete used for conventionally bored
be done in dedicated software, where tunnels, for cut and cover or immersed tube
The curve indicating the capacity of the joint the uncertain values are modelled as tunnels use ‘normal’ strength concrete (30Mpa
can be indicated for points of time during stochastic variables. This can be done with - 40Mpa). This type of concrete will begin to
and after the fire as the normal force and a full representation of the uncertain values lose its strength when its temperature exceeds
moment. as stochastic variables or as a more simple 380°C. Spalling may also occur under severe
sensitivity analysis, in which the uncertain fire exposure conditions in tunnels.
5.7.3 Capacity of elements in bending values can be varied between e.g. +/-1 or
two standard deviations or similar. In addition, the temperature of the steel
Sagging reinforcement is normally positioned reinforcement within the concrete (particularly
close to the heated surface of the concrete, The result of the fire resistance estimation the soffit) will begin to lose its strength when its
and the assessment of the structural integrity shall be given with the information of these temperature exceeds 250°C.
is governed by the reduced, remaining evaluations.
capacity of the reinforcement. Safety considerations, i.e. partial In order to prevent loss of strength and
coefficients and similar, will strongly stiffness (and potentially spalling) of tunnel
In case of spalling the reinforcement may influence the fire resistance. The best linings, measures should be taken to prevent
be exposed to the higher temperatures and estimate on the fire resistance will be rapid temperature rise of the structure or lining.
even become directly exposed to fire, and found by calculation with factors equal One of the most common solutions is to apply
consequently lose its strength much more to 1 and with central estimates of the an insulating coating (passive fire protective
rapidly. material characteristics. A best estimate coating). Other solutions exist such as addition
can make sense in connection with of polypropylene fibres but are not discussed
Hogging reinforcement will normally be risk analyses. However, for evaluation in details in this document. The advantage of
heated only marginally. The moment capacity of structural behaviour and structural passive fire protective coating is to not damage
can often be calculated by as an equivalent safety characteristic values and factors the concrete and optimize the repair work after
cross-section, where the concrete reduced according to the respective codes may be a fire (see paragraph 3.1.3).
to less than e.g. 60% of its original strength required. In the latter cases the estimated
is disregarded. time to failure will be a lower fractile. It This structural fire protection system may be
shall be taken into account in estimation either a proprietary spray applied material or a
For elements subjected to combined of characteristic values (and safety) which rigid or flexible board system fixed to the tunnel
bending and normal force, reference is part of the information has been found by lining.
made to the comments above concerning testing - and the uncertainties associated
capacity of segments. to the tests. The advantages of passive fire protection
systems are, amongst others:
5.7.4 Fire Resistance Moreover, such analysis is valid only under • Prevents rapid temperature rise in the
the assumption that concrete spalling and/ structure
When the relative curves indicating the or cracking will not occur. As stated in
• Prevents loss of strength and stiffness of the
capacities of the segment in terms of normal paragraph 3.1.4, concrete spalling cannot
load bearing concrete lining
force, moment, splitting and shear are not be predicted by numerical tools. If there is a
intersected by the load curve in the same risk of spalling, it needs to be assessed by • Prevents chemical degradation of the
units then the fire resistance has not been a fire test on the concrete. Fire resistance concrete
reached. A parametric extension of the fire analysis of concrete structure should then • If designed properly, the concrete is not
duration until the curves intersect can be take into account these experimental compromised
made. The duration of the fire, which results results. - Concrete spalling can be prevented
in intersection, represents the fire resistance. - Irreversible deflection can be prevented

The structural fire protection system must be the inside (fire exposed) and outside (non- Where masonry linings are used on walls
capable of limiting the temperature of both exposed) sides of the lining. These cracks separating two adjacent tubes where
the concrete and the steel reinforcement can obviously not be seen but repair works the adjacent tube is used as a means of
below their critical temperatures under the have to take them in consideration because escape, the use of an insulating coating
relevant fire exposure conditions (time- they can potentially cause significant mid/ may also be considered.
temperature curve). long term durability issues. Concerns have
also been expressed as to the load bearing 5.8.5 Gaskets
In immersed tube tunnels, consideration capacity of the structure predominantly
must also be given to the critical temperature with regard to its shear force capacity. The water tightness of segmental lining
of both immersion and segment joints in is generally ensured by use of gaskets.
order to maintain the structural stability of 5.8.2 Steel Linings In most cases the gaskets are placed
the tunnel under the relevant fire exposure near the extrados, but in some cases a
conditions. Steel linings will begin to lose their supplementing gasket has been placed
structural strength when its temperature also at the intrados. The gaskets have
In mechanized tunnelling, tunnels are exceeds 500°C - 550°C. If the mechanical no load bearing effect but a failure of the
usually lined with high strength pre-cast properties of the steel structure are to be gaskets will lead to more or less water
concrete seg-ments (55 - 90Mpa), which strictly preserved, it is therefore necessary inflow and subsequent repair. It will not
under fire exposure conditions, will begin to limit the critical temperature of steel be possible to replace the gaskets, so the
to spall at lower temperatures than cast linings below this temperature under the water tightness will have to be established
concrete. Experience has shown that relevant fire exposure conditions by the use by other means. It should be discussed
these types of concrete might start spalling of an insulating coating or lining in the same how much the tunnel should leak before it
at temperature levels well below 250 °C. If way as for concrete linings. A safe limiting is regarded as failed.
the choice is made to avoid any concrete temperature, also aiming at preventing
damage and reduce the repair constraints, irreversible deformations, application of It is beyond the scope of this note to
it is therefore necessary to prevent this type which makes further structural analyses describe the possible types of gaskets,
of concrete from exceeding temperatures superfluous, for steel linings is 350°C [5.5]. but in many cases the gaskets are made
of 180°C to 220°C after 120 minutes of fire from a rubber compound.
exposure under the relevant conditions in If the lining is constructed using segments,
order to preserve the full lining thickness consideration must be given to the critical The heating of the gasket depend both on
for the duration of the fire. However, during temperature of the tunnel joints and seals in the location and adjacent materials. In case
a fire test these temperatures are not the order to maintain the structural integrity of of a steel or cast iron segment the gaskets
only criteria to judge if the system failed or the tunnel under the relevant fire exposure will be heated more than in a concrete
not, and a “no spalling” criterion is also to conditions. segment. The calculation of the heating
be controlled. of the gaskets can be made together with
5.8.3 Cast Iron Linings the determination of temperatures in the
Pre-cast concrete roof beams may be segment (in this case in minimum a 2D
treated in the same way as pre-cast Cast iron linings may be treated in the model).
segments. However, beams are exposed to same way as steel linings with respect to
fire from multiple sides whereas segments their critical temperature. It may be relevant to ask the producer
or slabs are exposed from one side only. of the gaskets for specific material
Multiple side exposure of structural In ‘wet’ cast iron lined tunnels, consideration information. The following may serve as
elements (such as, but not limited to, must be given to the tunnel seals which a guideline for the characteristics of the
beams) causes elevated temperatures may have a lower critical temperature that gaskets subjected to fire:
at the corners and edges introducing the lining (for example, lead seals will begin •F
or even long duration exposure to
additional risk for early concrete spalling. to melt at approximately 300°C). about 150°C the appearance of the
rubber does not alter, the elasticity is
Research has been conducted on 5.8.4 Masonry Linings decreased, but in many cases it may
the effects of fire on unprotected serve its purpose also after the fire.
concrete linings which contained Masonry linings are generally considered This depends of course on the margin
polypropylene fibres. Reference is made to be fire resistant. However, under the of water pressure for which the gasket
to test reports 2009-Efectis-R0358 and most onerous fire exposure conditions, was designed. Deformations during the
2008-Efectis-R0146. In summary it has consideration may be given to the use of fire may in theory influence the long term
been concluded that post cooling cracking an insulating coating to prevent extensive sealing characteristics.
of the concrete lining will occur at both damage or collapse.

•F
or long term exposure to 250°C the The joint normally has a component of a 5.9.2 Steel Ceilings
rubber becomes brittle and loses all butyl rubber or similar, in form of e.g. an
elasticity. As it is probably subjected omega or gina joint. This material cannot Where used, steel suspended ceilings may
to high mechanical pressures from the withstand fire or heat, as mentioned in the be treated in the same way as concrete
segments during the fire, the gasket section on gaskets above. ceilings, particularly where the void is used
will most likely have lost its sealing as a mean of escape.
properties after the fire. A replacement For cut and cover tunnels the joints are
with another sealing medium will be often a less critical part of the structure. Since the suspended ceiling is unlikely to
necessary. For joints with rubber components the be fully loaded, a higher critical temperature
same comments as above apply. For than that of the lining may be considered.
Heat ageing and chemical attack may be a concrete-concrete or concrete-steel-
problem for the areas exposed. However, concrete construction joints, the fire Suspended ceilings constructed of cast
only a very small area will be exposed and resistance will not be critical. iron or masonry (if used) should meet the
hence the problem small. same requirements as those for concrete
5.9 Suspended Ceilings and steel.
The fire resistance of the gaskets is
dependent on the protection. If the 5.9.1 Concrete ceilings 5.9.3 Multiple sides fire exposures
gaskets are exposed directly to flames
they will be destroyed by fire. Loading on concrete suspended ceilings For example at the location of smoke
can be limited to the dead load of the inlets, the suspended (concrete) slab
5.8.6 Packers ceiling and pressure from the ventilation gets exposed to fire from both below
system. However, numerous tunnels use (the tunnel) and from the top (the smoke
Packers between the segment rings have the ventilation plenum as escape route extraction duct) because hot gasses are
the purpose of distributing the load during and means of egress, in which case these pulled into the duct. In the design stage
erection of the segments. The packers loads need to be accounted for. of such partitions it needs to be taken into
are often made of bituminous material or account that multiple sides fire exposure
of wood, and hence combustible. False ceilings and walls separating to a structure is different from single
•T
he bituminous packers may melt ventilation ducts from a tunnel should side exposure. For example, spalling of
and the melted material burns. It is have a minimum fire resistance (N0) where concrete will occur more rapidly and with
considered that this is a minor problem loss or collapse is prevented. much more explosive nature in the case of
as the packers will melt relatively slowly In other circumstances, where continuity multiple sided fire exposure.
and the combusted material will be of the suspended ceiling is essential,
small compared to the fire load in such as escape routes, higher levels of 5.9.4 Loads and structural behaviour
general. On the other hand the heating, fire resistance may be required. However,
softening and melting of the packers if such partitions are lost in a fire event, As already mentioned, the suspended
will to some degree relief the stresses the ventilation strategy along with the ceiling can form the ducts for the
originating from the thermal expansion tenable environment is lost too. Therefore ventilation, which is a key safety factor in
in the longitudinal direction. the protection of these partitions shall be case of fire. Despite this, in some cases,
considered. failure of the ceiling can be accepted just
•P
lywood packers will hardly burn due to
over the fire, but it has to be carefully
the limited surface; the relieving effect
Consideration should also be given to the investigated as the ventilation strategy will
from heat of the plywood packers will
interfaces between the suspended ceiling/ be endangered.
be small.
walls and the tunnel lining, especially if Reduction of strength in case of fire is
structural integrity is required in order to similar to the mechanisms mentioned for
5.8.7 Immersed tunnel Joints
maintain the stability of the tunnel and the lining. The structural particularities of
allow safe escape. the ceiling are:
For immersed tunnels it is particularly
important that joints are well designed he structural interface at the tunnel
•T
In voids above suspended ceilings or wall supports/ bearings or cast-in
and that they are watertight under all load
walls which are used as means of escape, reinforcement bars.
conditions. With respect to fire the joints
consideration should be given to provide
are normally placed at a position where •T
he supports are sensitive to shear
a maximum temperature of 60°C on the
they are protected against the heat. failure, movements resulting in loss of
upper (non-exposed) side of the structural
Otherwise they will have to be protected supports, and spalling / disintegration of
element.
by fire protection material. the concrete.

• The suspended ceilings will to some 5.11 Anchorages Italy), Tauern Tunnel (Austria), UPTUN –
extend (if it remains in place during the Runehamar Tunnel (Norway).
fire) protect the main structure from being
Internally in lining and other main structures,
subjected to the fire ; when the ventilation 5.12.2 Example of Circular Bored Tunnel
bond is necessary to obtain shear and
ducts are used for extraction of the hot (Shield Tunnel)
bending capacity. Similar for secondary
gasses, the main structure will also be
structural elements like suspended ceilings,
heated by the gas in the duct.
finally anchorage is necessary for fixation Figures 5.1 shows the Trans-Tokyo
of equipment like jet fans etc. inside the Bay Expressway Tunnel in Japan (TTB)
It is discussed whether a fire protection of a tunnel. standard cross-section and Figure 5.2
tunnel with suspended ceiling should be on shows the same of Elbe Tunnel of Germany,
the lower side of the ceiling or on the main respectively. On the one hand, the 4th
Large diameter anchors pose a risk, due
structure. As both structures have a specific Tunnel has an outer diameter of 13.75m,
to heat sink effects, locally increasing
function in a fire scenario, the answer can and an inner diameter of 12.35m (segment
the spalling sensitivity of the surrounding
be found according to the necessity to thickness of 70cm).
concrete.
preserve one or more functions in reference
to the safety strategy.
Cast-in item shall be considered due to
5.10 Supported Floors and Decks failure of their own function (anchorages for
installa-tions for jet fans for example may
fail and the fans will drop). Furthermore
Supported floors and decks will have a high the influence on the segments shall be
live load due to traffic movement. considered. The thermal expansion etc.
of the cast in item can be compared with
Structural integrity of the floor or deck must the concrete, in order to consider damage
be maintained where the void below is used to the structure from heating of the cast
for escape. It is also important that the in item. The cast in item can function as a
stability is maintained in order to allow the path for the heat into the concrete (for large
rescue and emergency services to perform items only).
their function.
On the other hand box-outs hollow parts
Figure 5.1 : Standard TTP Cross-section [5.6]
Propagation of fire underneath the and smaller voids can function as drains
supported floor or deck must be prevented for the migrating water and may reduce the
(particularly with the flow of burning spalling in this area.
hydrocarbon spills such as petrol). Anchorages for fire protection product are
discussed in section 5.12.6.
If there is another level of traffic below the
supported floor deck, consideration should 5.12 Mitigation Technology for
be given to maintaining the stability of the Tunnel Structures
deck under fire exposure.
5.12.1 Bored Tunnel (Horse Shoe Tunnel)
in stable rock conditions
The floor has normally one or two
intermediate supports and will span
between the supports and the tunnel For in-situ concrete lining and sprayed
wall. In connection with a fire it is of high concrete lining, it is verified through
importance that the rescue forces can use experimentation on actual tunnels that
the road (deck) when they enter the tunnel. horseshoe tunnels are generally not prone
The degree of damage of the floor may to severe damages in fires, water flood
also be one of the determining factors for and structure collapse. These cases are: Figure 5.2 : Standard Cross-section of No.4 Elbe Tunnel [5.7]
the decision of resuming the traffic after a EUREKA Repaaford Tunnel (Norway),
closure of the tunnel due to fire. Memorial Tunnel (USA), Nihonzaka Tunnel
(Japan), Mont Blanc Tunnel (France

When to compare the main characteristics

of these two tunnels :
• In TTB tunnel, there is a secondary
lining of 35cm thickness. Therefore, the
effective inner diameter of the tunnel has
become smaller than that of No. 4 Elbe
Tunnel.
• On the other hand, in No.4 Elbe tunnel,
without using a secondary lining, it is
planned to install fire resistant material,
and therefore, the effective space above
the carriageway is larger than that of TTI3
Tunnel.
Figure 5.3 : Cross section of Liefkenshoek Road Tunnel, Antwerp, Belgium
Both these tunnels are underwater
tunnels, that it is necessary to consider structure and the entire structure would be All the fittings used for the fixing of
the buoyancy of the tunnel bodies. In TTB influenced in case the duct is damaged. equipment to the structures should be
Tunnel, this secondary lining plays the role This is mainly the case for immersed tunnels considered in relation to the fire situation.
of counter weight suppressing buoyancy, in and tunnels of rectangular cross-section This means that use of the usual plastic
addition to the protective effect it provides with cut and cover construction method plugs and similar devices should be
on the segment itself. Therefore, it has (Figure 5.3). prohibited. This material will either soften
become possible to use the passage below or melt at high temperatures, resulting
the carriageway as escape route and an Ad (ii) In the following case, see Figure 5.4, in flames. Further research will have to
access path in emergencies. the suspended slab making up the lower demonstrate to what extent special plastic
part of the ventilation duct, is not a critical is suitable for fixing equipment inside the
On the other hand, it is necessary to adopt part of the load bearing structure. tunnel.
anti-buoyancy policies using passage
below the carriageway as a counter weight. Also steel plugs and anchors should be
As described in the above two categories checked for their behaviour during a fire
of examples, there exist two approaches situation. The tensile stress reduces at
for shield type road tunnels with large cross high temperatures. The calculation of these
sections. anchors should be based on the maximum
temperatures at which the anchoring
5.12.3 Mitigation measures for should be still properly working.
protection of ceilings in the
Liefkenshoek Road tunnel A distinction must be made between
major equipment, which should remain in
In road tunnels ceiling ventilation ducts are function during an emergency situation,
installed for transverse ventilation, semi and equipment whose functions may not
transverse ventilation and some kinds of be needed during or after a major disaster.
longitudinal ventilation lateral flow, semi
lateral flow, or partially in longitudinal flow The minimum requirement is that heavy
systems. equipment should not fall down when
Figure 5.4 : Typical cross section of horse shoe, transverse evacuating users or rescue personnel are
ventilation tunnel
There are two different cases in the structure in the tunnel. This means that no heavy
of this ventilation duct (i) the duct is a part of item must fall under 400°C-¬450°C
the tunnel’s structure itself, which basically 5.12.4 Mitigation measures for during the time necessary to fight fire (in a
receives stress and strain, and (ii) the duct protection of anchorages in traffic tunnel, such temperatures can produce a
is not a part of the structure in terms of live space radiation level of about 5 kW/M2 which is
loads. the maximum tolerable value for firemen).
Regarding to fire protection for suspending
Ad (i) This structure should be well heat structures and other systems attached to As aluminium loses its strength at a
resistant as the ceiling duct is a part of the the ceiling or walls, PIARC described below. temperature of approximately 550°C, it

is recommended to consider the use of thermal insulation, as the most important (E) Freezing and thawing resistance
aluminium materials in a tunnel critically. aspects of the passive fire protection Especially when the fire prevention
Alternative materials are steel and stainless system performance: materials are fixed in tunnels where
steel. temperature differentials year around are
Especially, jet fans located at tunnel ceiling (A) Installation strength severe, the protection material is subject
or sidewall for ventilation and smoke control Interior of traffic tunnels face severe to reciprocal effects of drying (thawing)
are one of the heaviest equipment. If the dynamic pressure differential due to passing and wetting (freezing) due to precipitation
temperature rise is sufficiently high, the fixing traffic and piston effect in ventilation ducts. outside the tunnel. Therefore, it is important
part at the top of ceilings for fans may be Numerous tunnels have had face pressure that freezing and thawing resistance in
impaired and equipment may fall down to differentials measured in the order of these materials is evaluated.
carriageway. magnitude of 2 kPa. On the other hand,
it is reported that rail tunnels experience (F) Water leakages and humidity
Therefore, this heavy equipment should approximate pressure differentials of 6 kPa With reference to point (E), fire protection
be mounted on the crossbeam which are when high speed trains pass the tunnel. materials in tunnels can get wet due to
fixed at side wall. Because of temperature Therefore, reciprocal differentials of this (minor) leaks in the structural concrete
distribution at side wall is slightly lower than nature should be withstood by the fire tunnel lining, which will cause a situation
temperature at ceiling (Figure 5.5). protection material including the fixation where the fire protection materials will get
materials and sub-frames. Test evidence fully saturated with water. High relative
by an authorised lab should be presented. humidity or water spray due to passing
vehicles (in case of wet road surface) will
(B) Prevention of secondary effects also cause water absorption of the fire
As tunnel temperature is extremely high, protection material.
fires, smokes and toxic gases may occur. It is reported that some water saturated
Hence it is necessary to use materials fire protection materials will not be able to
that are non-combustible (class A1) withstand a fire anymore. This is due to the
when tested to EN 13501. Also note the fact that the water evaporates (expansion
comments made in par 3.3.2, related ratio from water to steam is 1:1700) and
to non-suitable products for passive fire the vapour cannot escape through the
protection in tunnels. pores, creating pore pressure build up in
the material itself, causing a spalling effect
(C) Harmful effects of material and decomposition of the material, just like
It is necessary to assure that the material concrete will spall.
is free from harmful effects to humans Test evidence by an authorised lab
Figure 5.5 : Typical support system for ventilation fans should be made available, proving that
during tunnel construction. Materials
that are produced in production facilities, fire protection materials can meet the fire
having an NEN-ISO 14001 (environmental) protection requirements with less than 5%
Anchors will cause a heat sink into the
certificate are therefore preferred. Also (by weight) humidity and also when fully
concrete, potentially causing the concrete
note the comments made in par 3.3.2, saturated with water.
to spall. In such an event the fire protection
material is pushed from the concrete related to non-suitable products for
passive fire protection in tunnels. (G) Repair and Maintenance
lining, exposing the load bearing structure
to a fully developed tunnel fire. In such a Tunnel linings often get washed and
scenario the load bearing capacity of the (D) Constructability cleaned on a regular basis. The materials
tunnel lining is endangered. Anchor details As far as construction cost and quality used should be resistant to cleaning actions
like this need to be designed properly and assurance of fire prevention material are coming from brushes, water pressure and
heat sink effects have to be taken into concerned, it is necessary to minimise the use of mild detergents. Tunnel linings
account. loss in installation, control of rebound should also be easily repairable to avoid
water volume (for sprayed systems), downtime of the tunnel. In the aftermath
5.12.5 Protection of Structural consideration for adjustment of standard of a limited tunnel fire, the material should
Elements measurements in panel type fixation are be inspectable by the manufacturer to
exercised. Further, to prevent hindrance determine the level of degradation of the
When installing fire prevention material in to fire resistance qualities, construction material and the necessity to replace it, or
road tunnels, it is necessary to consider the standardisation is adopted. not.
following requirements, in addition to the

The tunnel owner might consider it (A) Major characteristics after the surface has been properly
necessary to inspect the tunnel lining, • Even in case of complicated cross prepared
for cracks and leakages, especially (but sections, construction is simple heir fire performance is very predictable
•T
not exclusively) near/around joints. This and often provides protection to the
•A
t narrow locations like pipe sleeves,
can have implications on the (partial) concrete substrate for periods of time in
construction is very appropriate
removability of protection systems, and/ excess of its intended fire rating.
or non-destructive inspection techniques •R
enovations are simple at places where
concrete is damaged after the concrete • Instances have been documented where
(e.g. acoustic). The choice for a proper
surface has been cleaned to create a vermiculite cements have prevented
fire protection system can seriously
proper bond concrete substrates from spalling under
be affected by these specifications.
extreme fire conditions for periods
Spot checking for (more than micro)
(B) Cases needing improvement of up to 24 hours, at which time the
cracks, leakages and deformations are
concrete/fire proofing interface has
recommended. • It is necessary to make adjustments if the
reached an equilibrium.
In any case, inspection of the tunnel adhesion of concrete lining varies due to
structures has to be done and recorded surface conditions. This may mean that
Panel / Board systems
before implementing fire protection. a mechanically fixed wire mesh, made
out of stainless steel or plastic coated
The most commonly used system to
5.12.6 Type of Fire Protection steel, has to be installed
protect concrete tunnel structures from
Materials • Since the finished surface is not flat, fire is the use of fire protection boards. The
cleaning is troublesome, and again if flat thickness of passive fire protection boards
As far as the fundamental concept of surfaces are needed for certain cases, is engineered to suit the fire requirements
fire resistant policies are concerned, then special considerations should me of the project. Such boards always need to
considering the temperature rise of made be mechanically attached to the concrete
between 1,000ºC – 1,350ºC within a
• To ensure a homogenous finish and lining.
short period by vehicle fire, it is common
resistive effects at required level, then
to allow for temperatures as high as the
it is necessary to maintain a consistent Specific properties of board systems for the
permissible temperature of fire resistant
level in technical standards fire protection of concrete structures are
materials of around 1,350°C, in case of
• Thickness control of the applied material listed below:
concrete and reinforcements. The melting
temperature of fire protection materials must be carefully controlled. •B
oard materials can easily be checked for
shall be greater than 1350°C. thickness and thus the application can be
(C) Basic performance guaranteed to meet with the specifications
Spray Type 1. Organic coatings [5.8] as per the tested constructions.
•O
rganic coatings produce toxic fumes eing mechanically fixed, board systems
•B
The application process of Spray type and dense smoke under fire conditions, can cope with the dynamic loads from
systems involves a number of stages. which make them unsuitable for use in passing vehicles (pressure / de-pressure
tunnels where life safety is an important loads)
First, the concrete surface needs to be consideration. • In general, cement based board materials
cleaned to remove dirt, dust, demoulding are unaffected by water ingress in the
oil etc. A wire mesh is then mechanically 2. Inorganic coatings [5.9] tunnel and combustion gasses.
fixed to the concrete which will serve as a
• Inorganic coatings are produced as •B
oard systems in general require no
retaining system, should the spray mortar
factory controlled cement/ vermiculite maintenance. In the case that the concrete
lose its bond to the concrete surface,
premixes, which are sprayed, applied substrate needs to be inspected, when
during the lifetime of the tunnel. Vibrations
directly to the internal surfaces of the properly designed and installed, the
and sagging of the tunnel may cause
tunnel lining. boards can be removed and reinstated,
spray system to detach from the concrete
•V
ermiculite cements are essentially thus maintaining the fire protection layer
lining, hence the use of wire mesh. A
inorganic materials and therefore will not at all times.
keycoat provides a better adhesion to the
concrete surface. The sprayed on material burn, produce no smoke or toxic fumes oard materials are produced in factories,
•B
can be left as a sprayed finish or can be and are certified non combustible. ensuring the quality of the material and
trowelled to a smooth finish, to receive an •T
hey can be spray applied quickly to the a correct composition. Most board
optional paint system as a final finish. tunnel surface and if damaged, can be materials are produced in NEN-ISO
Fundamental performance of the spray easily repaired (by hand in small areas) 9001 and NEN-ISO 14001 certificated
type fire protection material is as follows : factories.

Two methods of installing boards systems •T

he anchors should be suitable for use In case of fire, the temperature rise in the
can be defined : in the tension zone of concrete (cracked concrete will lead to damages. Repair
(1) Lost shuttering method: the boards con-crete) works will be necessary to re-establish the
will be installed in the form-work, on top •T
he anchors should be suitable for use initial structural performance if needed.
of the load bearing plywood. Screws are under dynamic loads
partly inserted into the fire rated boards 5.13 Current Production Lists
to create a bond between the panel and 5.12.7 Development of Fire Resistant of Fire Resistant Materials
the concrete. After the reinforcement is Concrete
installed the concrete can be poured. After Many materials are currently manufactured
the extraction of the form-work the fire for the protection of structural elements in a
With the objective of protecting the tunnel
rated boards will stick to the concrete thus tunnel structure. Many materials are similar
structure from high temperatures in case
creating the fire protective layer. The screws to those used for the traditional protection
of fires, fire resistant concrete has been
are anchored into the concrete creating of buildings in providing a certain fire rating
developed, having upgraded the quality of
a firm bond. Due to the very low labour as required by local codes and ordinances.
concrete. In one such method, it has been
costs of this system, it’s intensively used in These local regulations should be consulted
attempted to increase the fire resistance of
cut & cover and immersed tunnels, rather in the selection and type of material to
concrete by mixing chemical compounds
than having to apply a fire protection layer be used for protection of the structural
to those existing in concrete. For example
afterwards. elements of a tunnel system.
several tunnels have been built in Europe
(2) Post installation: in both new and these last 10 years adding polypropylene
existing tunnels the post installation fibres to the concrete. As a result, these
method can be chosen. The boards can be tunnels have been assessed to be fire
installed, either directly to the concrete or resistant without the use of fire protection.
on a sub-frame. We can note that :
• Calcium silicate boards are suitable for
installation on flat concrete substrates.
MATERIAL TYPE MATERIAL TYPE CONST ATTACHMENT
• Calcium silicate aluminate boards
may be installed on curved concrete Anchor bolts for post fixed
substrates and have an improved thermal Calcium silicate board Panel Pre-manufactured panel
system
performance over conventional board Calcium silicate board Panel Pre-manufactured panel
Screws for lost formwork
systems. system
Anchor bolts for post fixed
Calcium silicate aluminate system, either flat or post
Panel Pre-manufactured panel
The installation of fire protection materials board curved (flexed) into radius
of tunnel
should be done with anchors having the Screws for lost formwork
Calcium silicate aluminate
following properties: Panel Pre-manufactured panel
board system
• The diameter should be limited to Cement based spray-on
Spray mortar, cementitious
Applied onto flat or curved
vermiculite or perlite based Spray
maximum M6, in order to reduce the heat systems
Cement / additives / fibres
tunnel lining
sink effect through the steel anchor into
the concrete. It is reported that thicker Table 5.1: Typical fire protection materials for tunnels
anchors can create a local spalling effect
of the concrete; this local effect is only
temporary because the spalling will
spread all over the surface once a small
part of the concrete is directly exposed
to fire.
• The use of stainless steel anchors is
recommended. Types that can be used
are A4, 316, 1.4401 and 1.4571. In some
countries even higher requirements are
applicable by code.
• If necessary a washer must be used
to avoid a pull through effect when the
system is exposed to dynamic loads

6 >> Summary and recommendations
The goal of the Working Group is to provide those standards by providing the providing careful consideration is made
guidelines and recommendations to ensure following recommendations to be used as to the tunnel usage, and types of
fire resistant tunnels contributing to the as guidelines for the protection of road materials that are transported within the
global safety level, reduce non-operational tunnels. tunnels. These guidelines are not intended
time and economic loss after a fire and to •T
hese recommendations were developed for passenger rail tunnels, which will be
prevent the catastrophic collapse of the by an extensive review of codes, covered in a parallel document to be
structure. regulations, ordinances, research and developed by the Working Group in the
• The Working Group has also case studies of tunnel fires. These future.
acknowledged that many of the guidelines are intended to provide criteria
member nations have already in place for the protection of existing as well as In agreement with PIARC, the Working
strict guidelines for the protection new road (highway) tunnel in construction. Group recommends that the design criteria
of underground structures, and the The information provided herein, may presented in table 6.1 be used according
Working Group hopes to enhance be applicable to other types of tunnels, to the type of structure and traffic in the
tunnel.
Main Structure Secondary Structures4
Immersed or
Tunnel in unstable Tunnel in stable Emergency Emergency exits
Traffic Type under/inside Cut & Cover Air Ducts5 Shelters6
ground ground exits to open air to other tube
superstructure
Cars/Vans 2 2
Trucks/Tankers RWS/ HCinc RWS/ HCinc RWS/ HCinc RWS/ HCinc ISO ISO RWS/ HCinc RWS/ HCinc
/HGV 120 min1 120 min1 120 min1 120 min1 120 min1 30 min 120 min 120 min7
Table 6.1: Joint recommendations of PIARC and ITA

Notes :
1
180 min maybe required for very heavy traffic of trucks carrying combustible goods
2
requirements:
3
requirements:
• RWS/HCinc 120 min if strong protection is required because of property (e.g. tunnel under a building) or large influence on road network
4
Other secondary structures should be defined on a project basis
5
In case of transverse ventilation
6
Shelters should be connected to the open air
7
A longer time may be used if there is a very heavy traffic of trucks carrying combustible goods and the evacuation from the shelters is not possible within 120 min
The document provides guidelines for the walls, cast-in-place concrete etc) will these thresholds, the concrete or steel
thermal protection of the various elements decrease once the concrete surface elements will be damaged with visible
of a tunnel. This is due to the fact that there temperature reaches 380oC ; for steel cracks, but also invisible ones because of
are so many tunnel sub-systems and types structural elements, this threshold is their seize or their location in the structural
of interior structural elements, it would be set around 550°C ; the corresponding element; repair works will be necessary to
impossible to identify each and every type impacts in terms of stiffness and resistance replace the damaged elements or stretches
and provide a specific fire protection system reduction have necessary to be taken into and recover the original mechanical
for them. account in the design and especially the performances;
dimensioning of the structures in the vehicle • With rapid heat rise, in accordance to
Among the main conclusions, we can fire scenarios; this document provides a few the standard time-temperature curves
highlight the following: elements and references for the evaluation described in this document, the concrete
•T
he mechanical resistance of structural of these impacts; can spall; this phenomenon is dependant
concrete elements (wall, ceilings, partition • After a fire leading to temperature above to many parameters and very difficult to

6 >> Summary and recommendations
estimate in common practice; because of its

possible impact on the structural resistance,
it has to be taken into account with a specific
evaluation, most likely a representative full
scale fire test;
• In sensitive tunnels, i.e. tunnels in unstable
ground conditions, or immersed tunnels,
the choice can be made, as a precautionary
measure, to prevent any structural resistance
loss, leading to a thermal protection of the
structural elements; the objective is therefore
to limit the concrete surface temperature to
380°C (or 180-220 C, for concrete types
more sensitive to spalling); such a choice
can also be made in general for all kinds of
tunnels, to mitigate the risk of a long period
of tunnel closure for repair after a fire.
• The documents presents the different
types and characteristics of fire protection
materials ; it also provides information about
their installation;
• More sophisticated assessment methods
could be used to develop more economical
solutions taking into account other factors
influencing like temperature gradients,
heating rates, structural load levels,..;
performance-based approaches can be of
great interest in this objective.
The information provided here is intended

to act as a guideline and is for informational
purposes. Specific measures for the protection
are site specific and must be designed on an
individual basis.
Any inquiries in regard to the information

presented here should be sent to the
International Tunnel Association Working
Group 6 via its web site at www.ita-aites.org
or to the Animateur of the Working Group
Mr. Henry Russell c/o Mott MacDonald at
Hen-ry.Russell@mottMac.com, or to the
vice-Animateur Mr. René van den Bosch c/o
Promat International at
rene.vandenbosch@etexgroup.com.

7 >> Bibliography
7.1 References for section 1
[1.1] Ingason H, Lönnermark A, Li YZ (2011) Runehamar Tunnel Fire Tests. SP Technicial Research Institute, SP Report 2011:55
[2.1] World Road Association (PIARC), Fire and Smoke control in Road Tunnels, Paris, 1999, 2007.
[2.2] World Road Association (PIARC), Systems and Equipment for Fire and Smoke Control in Road Tunnels, Paris 2004
[2.3] World Road Association (PIARC), Road tunnels: Operational strategies for emergency ventilation, 2011
[2.4] World Road Association (PIARC), Fixed fire fighting systems in road tunnels: current practices and recommendations, 2016
[2.5] Li YZ, Ingason H, Maximum Temperature beneath Ceiling in a Tunnel Fire. Technical Research Board of Sweden SP Report 2010
[2.6] World Road Association (PIARC), PIARC Proposal on the Design Criteria for Re-sistance to Fire for Road Tunnel Structures, Paris,
Revision 1 – February 2002
[2.7] Dutch Safety Board, Fire in a CNG bus, Wassenaar 29 October 2012, November 2013
[2.8] Li YZ, Lei B, Ingason H (2011), The maximum temperature of buoyancy-driven smoke flow beneath the ceiling in tunnel fires. Fire
Safety Journal 46 (4):204-210
[2.9] Li YZ, Ingason H (2012), The maximum ceiling gas temperature in a large tunnel fire. Fire Safety Journal 48:38-48
[3.1] Eurocode 1, EN 1991-1-2, “Actions on structures – Part 1-2: general actions – Actions on structures exposed to fire”.
[3.2] Eurocode 2, EN 1992-1-2, “Design of concrete structures. General rules. Structural fire design”.
[3.3] Carlsaw, H.S. and Jaeger, J.C., “Conduction of heat in solids”, Oxford University Press, 1959.
[3.4] Fire Protection in Tunnels and Underground Spaces, A Glerum, TU Delft (distributed at the ITA WG6 meeting in Oslo 1999).
[3.5] Phan, L. T. Fire Performance of high strength concrete: A report of the state-of-the-art. Res. Rep. NISTIR 5934, Gaithersburg, Md,
1997.
[3.6] Both, C. van de Haar, G.L. Tan, G. Wolsink, G. Evaluation of Passive Fire Protection Measures for Concrete Tunnel Linings. Tunnel
Fires and Escape from Tunnels, International Conference Lyon, May 1999. Both, C. Varley N. Fire protection of concrete lining in
tunnels, Concrete, May 1999.
[3.7] Steinert C., «Brandverhalten von Tunnelauskleidungen aus Spritzbeton mit Faserzusatz, Bauingenieur Bd 73 (1998), Nr 1 January
1998.
[3.8] Fire resistance of Tunnel Linings, Barry, I. E. Blake, A. Cafco Europe Group (distributed at the ITA WG6 meeting in Oslo 1999).
[3.9] International RILEM Conference on Materials, Systems and Structures in Civil Engineering Conference segment on Moisture in
Materials and Structures, 22-24 August 2016, Magnesium-oxide boards cause moisture damage inside facades in new Danish
buildings, Technical University of Denmark, Lyngby, Denmark
[3.10] CORE Project Consulting Independent Engineering Review of MgO Boards in Tunnel lining
[3.11] Journal of Chemical and Pharmaceutical Research 2014, The application review of magnesium oxychloride cement
[5.1] Technical instruction relating to safety measures in new road tunnels (design and operation), published as appendix 2 to the French
inter-ministry circular No. 2000-63 of 25 August 2000 concerning safety in the tunnels of the national highways network (the circular
and its appendix 1 are no longer in force, only appendix 2 is still applicable)
[5.2] Fire test procedure Ministry of Transport RWS/Efectis 2008-Efectis-R0695
[5.3] NFPA 502, Standard for Road Tunnels, Bridges, and Other Limited Access Highways, 2017
[5.4] CETU, Guidelines on Passive fire protection systems, March 2017
[5.5] Eurocode 3, EN 1993-1.2, 2005
[5.6] Trans Tokyo Bay Highway Kawasaki Tunnel Ukishima North (Phase 1), Trans Tokyo Bay Highway Kawasaki Tunnel Ukishima North,
Kumagaigumi, Hazama & JDC Joint Venture.
[5.7] DIE. 4. RÖHRE ELBTUNNEL, Bundesrepublik Deutschland, Bundesministerium für Verkehr, Freie und Hansestadt Hamburg
Baubehörde.
[5.8] Project 145.1 Fires in Transport Tunnels Report on Full Scale Tests, EUREKA; EU 499: FIRE TUN.
[5.9] Report concerning on an investigation into the behaviour of a concrete slab protected by a coating of PROMATECH H Plates of 25mm
in thickness upon heating according to a curve prescribed by the «Rijkswaterstaat» in the Netherlands, TNO Report B 85 191(E),
May 1985.

7 >> Bibliography
7.5 Some references to go further…
- K. Kordina, C. Meyer-Ottens, Beton Brandschutz Handbuch, Beton-Verlag GmbH, Düsseldorf, 1981

- C. Meyer-Ottens, Zur Frage der Abplatzungen an Betonbauteilen aus Normalbeton bei Brandbeanspruchung, Dissertation, TU
Braunschweig, 1972
- C. Meyer-Ottens, Abplatzversuche an Prüfkörpern aus Beton, Stahlbeton und Spannbeton bei verschiedenen Temperaturbeanspruchungen,
TU Braunschweig, 1972
- C. Meyer-Ottens, Zur Frage der Abplatzungen an Betonbauteilen aus Beton bei Brandbeanspruchung, TU Braunschweig, 1972
- P. J. E. Sullivan, The Effect of Temperature on Concrete, Developments in Concrete Technology - 1” London, 1979
- K. D. Hertz, Betonkonstruktioners brandtekniske egenskaber, Rapport nr. 140, Instituttet for Husbygning Danmarks Tekniske Højskole,
1980
- K. D. Hertz “Design of fire exposed concrete structures” Report No. 160, Institute of Building Design Technical University of Denmark,
1981
- K. D. Hertz “Eksplosion og reststyrke af varmepåvirket beton” Rapport nr. 162, Instituttet for Husbygning Danmarks Tekniske Højskole,
1982
- J. W. Dougill “Materials Dominated Aspects of Design of Structural Fire Resistance of Concrete Structures” ACI SP-80, 1983
- W. J. Copier “The Spalling of Normal Weight and Light Weight Concrete Exposed to Fire” ACI SP-80, 1983
- T. Z. Harmathy “Effect of Moisture on the Fire Endurance of Building Elements” ASTM Special Technical Publ. No. 385
- P. J. E. Sullivan, A. A. A. Zaman “Explosive Spalling of Concrete exposed to high Temperatures” Structural Mechanics in Reactor
Technology, 1971
- G. L. England, T. J. Sharp “Migration of Moisture and Pore Pressures in heated Concrete” Structural Mechanics in Reactor Technology,
1971
- FIP/CEB “Report on Methods of Assessment of the Fire Resistance of Concrete Structural Members” , 1978’’
- Z. P. Bazant, W. Thouguthai “Pore pressure in heated concrete walls: theoretical prediction” Magazine of Concrete Research, 1979
- National Bureau of Standards “Building Research Translation, The Behaviour of Concrete Structures in Fire - A Method for Prediction
by Calculation” NBS Technical Note 710-10, 1978
- National Bureau of Standards “Fire Effects on Reinforced Concrete Members” NBS Technical Note 985, 1979
- Y. Anderberg, O. Pettersson “Brandteknisk dimensionering av Betong Konstruktioner - Del 1”, 1992
- Y. Anderberg “Brandteknisk dimensionering av Betong Konstruktioner - temperaturbilagan”, 1992
- Y. Anderberg “Fire-exposed hyperstatic concrete structures - an experimental and theoretical study” Division of Structural Mechanics
and Concrete Construction, Bulletin 55. Lund University of Technology, 1976
- Y. Anderberg, S. Thelandersson “Stress and Deformation Characteristics of Concrete at High Temperatures” Division of Structural
Mechanics and Concrete Construction, Bulletin 54. Lund University of Technolohy, 1976
- F. P. Smith “Concrete Spalling: Controlled Fire Tests and Review” Fire and Arson Investigator, 1993
- B. Beland “Spalling of Concrete” Fire and Arson Investigator, 1993
- L. J. Parrott “Moisture conditioning and transport properties of concrete test specimens” Materials and Structures, 1994
- K. Boam, D. Cropper “Midlands Links Motorway Viaducts. Rehabilitation of a Fire damaged structure” Proc. Inst. Civ. Engrs. Structs
& Bldgs, 1994
- C. R. Midkiff Jr. “Spalling of Concrete as an indicator of Arson” Fire and arson Investigator, 1993
- C. R. Midkiff Jr. “Spalling - On the trail of Arson” Fire and Arson Investigator, 1993
- N. Andersen “Betonkonstruktioner - Brandteknisk klassifikation ved beregning og prøvning”, 1986
- T. H. Rasmussen, E. Yde “Beton og Brand” Beton Teknik, 1986
- M. S. Abrams, A. H. Gustaferro “Fire Endurance of Concrete Slabs As Influenced by Thickness, Aggregate Type and Moisture”
Development Laboratories, 1968
- H. Saito “Explosive Spalling of Prestressed Concrete in Fire” Spannbeton. Ergebnisse einer Tagnung der F.I.P in Braunschweig, 1965
- C. D. Roberts “Misconceotions About Fire Investigation / Part III” The National Fire & Arson Report, 1989
- J. L. Lentini “A Documented Case of Accelerant Induced Concrete Spalling”
- CIRIA “Fire Tests on ribbed concrete slabs” Technical Note 131, 1987
- T. T. Lie “Fire and Buildings” Architectural Science Series, Applied Science
- STUVO-report nr. 12 “Discussion on Fire Resistance: Lightweight concrete and spalling” Amsterdam, 1972
- K. Kordina “Labor brandversuche Fahrnan - Teil des FE-Vorhabens: Brandschutz in Verkehrstunnelanlagen” T U
Braunschweig, 1994
- B. F. V. Messerschmidt “Eksplosion og afskalning af brandpåvirket beton”, 12 points specialkursus Instituttet for Husbygning, DTU,
1994

7 >> Bibliography
- K. D. Hertz “Danish Investigations on Silica Fume Concretes at Elevated Tempera-tures” ACI Materials Journal, 1992
- Behaviour of concrete at high temperatures, Schneider, U. Deutscher Ausschuss für Stahlbeton, Heft 337, Verlag Wilhelm Ernst u.
Sohn, Berlin 1982.
- Concrete at high temperatures: Material properties and mathematical models, J. Eng. Mech. Div. ASCE 104, 1996.
- Ulm, F-J, Coussy, O. Bazant, Z P. The Chunnel Fire I: Chemoplastic Softening in Rapidly Heated Concrete, Journal of Engineering
Mechanics, Vol. 125, No. 3, March 1999,
- Ulm, F-J, Acker, P. Levy, M. The Chunnel Fire II: Analysis of Concrete Damages.
- Fasseu, P. and Kittel G. The trans-channel tunnel fire: Material tests. Rep. No. 96.6002532, March 1997; Fasseu, P. Eurotunnel Fire.
Analysis of the effect of fire on the concrete quality. Rep. No. 96.6002532, April 1997. Laborratoire Regional des Ponts et Chasseés
de Lille, France (in French).
- Cast Iron Physical and Engineering Properties, Angus, 2.ed 1976.
- Kordina, K. Fires in Transport Tunnels. Report on full scale tests. EUREKA EU499 FIRE TUN project. IMBMB Braunschweig, 1997.
- Fire Protection in Traffic Tunnels - Findings from Large-Scale Tests within the EUREKA Project EU 499 Firetun, Final Technical Report,
1995.
- Information on Brite project BE1158, HITECO and Fireexpo / Hitecosa distributed at a Fire Workshop Copenhagen 25.03.99.
- Y. Ota, K. Horiuchi, S. Date, «A Study on Thermal Analysis For Tunnel Structures in The Incidence of Fire», 4th International Conference
onSafety inRoadandRail Tunnels; Spain, April 2001.
- «Safety Aspects and Standards of Japanese Vehicular Tunnels», Fire Protection in Traffic Tunnels, Dresden, September 1995.
- A.G. Bendelius, A.S. Caserta, the Memorial Tunnel Fire Ventilation Test Program: Phase 111, Development of a Tunnel Specific CFD
Code, Report of Additional Session WG6: Fire and Smoke Control, PIARC M, World Road Congress, Kuala Lumpur, October 1999.
- Mont Blanc Fire Report finds coordination lacking, CNN Interactive July 7, 1999.
- KLEIN ZEITUNG, 30 Mai 1999.
- Trans Tokyo Bay Highway Kawasaki Tunnel Ukishima North (Phase 1), Trans Tokyo Bay Highway Kawasaki Tunnel Ukishima
North, Kumagaigumi, Hazama & JDC Joint Venture.
- RWS,TNO; Fire Protection in Tunnels, Document No. GT 98036 98 CVB RI 161, 26 November 1998.
- N.G. Zoldners, «Effect of High Temperature on Concrete Incorporating Different Segregates», Mining Branch Research Report, R64,
Department of Mines and Technical Surveys Ottawa.
- G.L. Tan, «Fire Protection in Tunnels Open to Hazardous Goods Transport», ECOLE NATIONALE DES PONTS ET CHAUSSEES, 10
December 1997.
- TUNNEL 7/99, STUVA, November 1999.
- Ian E. Bary, A. Blake, «Fire Resistance of Tunnel Linings», ITC Conference on Reducing Risk in Tunnel Design and Construction, Basel,
Switzerland, 7 8 December 1998.
- Ministry of Phblic works and Transport Flanders Region, The Liefkenshoek Tunnel
- Li YZ, Ingason H (2010) Maximum Temperature beneath Ceiling in a Tunnel Fire.
- SP Report 2010:51, SP Technical Research Institute of Sweden, Spalling of concrete and Fire Protection of Concrete Structures

>> Appendix 1 - road tunnel fire history
Fire Accident’s in the World’s Road Tunnels
consequencies
Vehicle Where Most Possible Duration of Consequences Structures and
year Tunnel Length Location Country Damaged Vehicles
Fire Occurred Cause of Fire Fire People Installations
Lorry with 66 Injured
Holland New York Load falling off lorry 10 lorries Serious Damage
1949 11 tons of 4h smoke
2,550 m USA Explosion 13 cars Over 200 m
carbondisulfide inhalation
Mont Blanc
1974 France-Italy Lorry Motor 15 min 1 injured
11,600 m
Crossing Lorry with drums of
Paris 12 light injuries Serious damage
1976 BP-A6 16 tons polyester High Speed 1h 1 lorry
France (smoke) over 150 m
430 m film
Velsen Velsen 4 lorries Front-rear 5 dead 4 lorries Serious damage
1978 1 h 20min
770 m Nederland 2 cars Collision 5 injured 2 cars Over 30 m
Nihonzaka Shitzuoka 4 lorries Front-rear 7 dead 127 lorries Serious Damage
1979 159 h
2,045 m Japan 2 cars Collision 1 injured 46 cars Over1,100 m
1 truck with 3600
Kajiwara Collision with side wall 1 truck, 4t Serious Damage
1980 Japan litres of paint in 200 n/a 1 dead
740 m and overturning 1 truck 10t Over 280 m
cans
1 car, 1 coach, 1 3 lorries
Caldecott Oakland 7 dead Serious Damage
1982 lorry with 33000 Front-rear collision 2 h 40min 1 coach
1,028 m USA 2 injured Over 580 m
litres of petrol 4 cars
Mazar-e-Sharif- Soviet Military
Salng Unknown, probably
1982 Kabul column. At least one n/a >200 dead n/a n/a
2,700 m mine ex-plosion
Afghanistan petrol truck
Gênes
Pecorila Galleria 9 dead
1983 Savone Lorry with fish Front-rear colli-sion n/a 10 cars Little Damage
662 m 22 injuries
Italy
L’Arme Nice Braking after 3 dead 1 lorry Equipment
1986 Lorry with trailer n/a
1,105 m France high speed 5 injured 4 cars destroyed
Gumefens Berne Front-rear 2 lorries
1987 1 lorry 2h 2 dead Slight damage
343 m Switzerland Collision 1 van
RØldal Røldal VW transporter
1990 n/a 50 min 1 injured n/a Little damage
4,656 m Norway With trailer
Mont Blanc Lorry with 20 Equipment
1990 France-Italy Motor n/a 2 injured 1 lorry
11,600 m tons of cotton destroyed
Serra Ripoli Bologne-Florence 1 car+lorry with 4 dead 5 lorries
1993 Collision 2 h 30min Little damage
442 m Italy rolls of paper 4 injured 11 cars
Hovden HØyanger Motor cycle 5 injured in the 1 motorcycle 111 m insulation
1993 Front-rear colli-sion 1h
1,290 m Norvay 2 cars collision 2 cars material destroyed
Huguenot Bus with 45 1 dead
1994 South-Afrika Electrical fault 1h 1 coach Serious damage
3,914 m pas-sengers 28 injured
1 lorry
Pfander 3 dead in the collision
1995 Austria Lorry with trailer Collision 1h 1 van Serious damage
6,719 m 4 injured
1 car
1 tanker Serious damage,
Isola Delle Femmine Palermo 1 tanker with liquid 5 dead
1996 Front-rear colli-sion n/a 1 bus tunnel closed for 2.5
148 m Italy gas + 1 little bus 20 injured
18 cars days
23 lorries
Serious damage,
1999 Mont Blanc Lorry with flour and Oil leakage 10 cars
France-Italy n/a 39 dead tunnel reopens
14 July 11,600 m margarine Motor 1 motorcycle
22.12.2001
2 fire engines
A10 Salz Salzburg-
Front-rear collision 12 dead 14 lorries
1999 Tauern 6,401 m Spittal Lorry with paint n/a Serious damage
4 cars and 2 lorries 49 injured 26 cars
Austria
The trailer truck that Front-rear colli-sion Serious damage.
E134 Dram-men- 1 lorry
caused the multiple A trailer- truck pushed a NOK 1 mill.
Haugesund 6 cars
2000 Seljestad 1,272 m collision had a diesel car into 4 cars that had 45 min 6 injured Tunnel closed for 1
Norway 1 motorcycle
fire in the engine stopped behind another ½ days
room before collision truck
A 32 Torino-
2001 Prapontin Romanian truck, 19 injured Closed until 6 June
Bardonecchia Mechanical problem n/a n/a
28 may 4,409 m loaded with beets by smoke in westerly direction
Italy
Gleinalm A 9 near Graz Front collision 5 dead
2001 Car n/a n/a n/a
8,320 m Austria Lorry-car 4 injured
13 lorries
St. Gotthard A2 Front collision Serious damage
2001 Lorry 2 days 11 dead 4 vans
16,918 m Switzerland Lorry-car Closed for 2 months
6 cars

>> Appendix 1 - road tunnel fire history
Sources :
CETU:
Incendies de tunnels routiers dans le monde, ayant occasionné des victimes ou dégâts importants, sans matière dangereuse impliquée
PIARC: XVIIIth World Road Congress, Brussels 13-19 sept. 1987, Technical Committee, Report No 5 Road Tunnels
[2.1] PIARC: Fire and Smoke Control in Road Tunnels, 05.05.B – 1999
ADAC: European Tunnel test 2001, April 26, 2001
FHWA-RD-83-032 Prevention and Control of Highway Tunnel Fires (appendix)
Longtunnel.com

>> Appendix 2 - Li and ingason model
The model developed by Li and Ingason [2.7, 2.8] is a simple and robust empirical model based on theoretical approach and a large amount
of large scale and model scale data. The maximum temperature beneath the ceiling in a tunnel fire is independent of the ventilation velocity if
the ventilation velocity across the fire source is very low compared to the heat release rate, and the maximum temperature is just dependent
on the heat release rate, however, it approaches to a constant if the part of the flame volume containing the combustion zone is present at
the tunnel ceiling. In other words, if V’ ≤ 0.19 (Region I), the maximum excess temperature can be expressed as :
The dimensionless ventilation velocity:
The dimensionless plume temperature:
The dimensionless position along the trajectory:
The characteristic plume velocity w* is defined as :
where bfo is radius of fire source (m), u is plume velocity (m/s), uo is wind velocity (m/s), s is trajectory (m), b is radius of fire plume at a given
position (m), g is the acceleration of gravity (m/s2), Q is heat release rate (kW), ρo is ambient density (kg/m3), cp is heat capacity (kJ/kgK), To
is ambient temperature (K), θ is the angle between plume axis and horizontal axis (°), α is tangential entrainment coefficient and β is normal
entrainment coefficient.
If the ventilation velocity across the fire source gets larger, the maximum excess temperature beneath the ceiling depends on both the heat
release rate and the ventilation velocity, however, it also approaches to a constant if the combustion zone is present at the tunnel ceiling. How
to determine which constant to use will be discussed in chapter 3. In other words, if V’ >0.19 (Region II), the maximum excess temperature
can be expressed as :
Example Scenario :
An HGV vehicle (Heavy Goods Vehicle trailer) is assumed to start to burn. The ignition can be related to the fact that one vehicle starts to
burn (a tire fire or engine problems). The fire can also be regarded as a fire that starts due to a collision of two HGVs. The tunnel height at its
maximum is 6 m and the width is 12 m at its widest part. The height from the bottom of the fire load up to the ceiling is 4.8 m. The radius of
the fuel in the vehicle, bfo, is 4 m. The fire is assumed to be fuel controlled. The ambient conditions are 10oC. This fire is expected to follow
the HC curve. Use the equations (1-3) to calculate the corresponding heat release rate curve and also if the fire duration is assumed to be
120 minutes, what is the corresponding fire load. The longitudinal ventilation is assumed to be 3 m/s. The Hydrocarbon Curve (HC) [3] can
be expressed as ΔT(t)=1080(1-0.325e-0.167t-0.675e-2.5t), where t is the time (min).
Solution : Firstly we assume V′ > 0.19 since Q(t) is unknown. This means that Equation (3) is used here. Then we calculate V′ using Qmax,
which turns out to be greater than 0.19.This means at any time in this scenario, V′ >0.19 is fulfilled, which verifies the first assumption. For
data with a temperature over 1100oC, it is difficult to determine in which region it lies. However, it is clear that the calculated heat release rate
is the minimum value required to obtain such a high temperature.
If we integrate the curve we find the total energy to be 496 GJ. This would correspond to at least two HGVs. A reasonable conclusion is
that this tunnel would be designed to a fire that can resist a collision between two HGVs, and the fire can be very intense for up to 120
minutes. This shows the practical implication of using Equations (1) and (3). These equations become a key to relate information between the
standardized time-temperature curves and the fire load in terms of the heat release rate and total energy found in the fire load.

Edited by: 100
Jan Brekelmans BSc (TNO Building and Construction
on
n Rese
Research,
earch, Cen
Centre for Fire Research)
200
René
é van den Bosch BSc (Promat, Tunnel Fire
re
e Protection)
P te o
300
400
Summary of Large Scale Fire Tests in the
500
y,
EHAMAR i Norway
600
gram
ram
n with the UPTUN Research Progra
ra
gra
r
gr
el in
Tunnel
EHAM
RUNEHAM
UNE
UN
RU
R EHAM
RUNEHA
RUNE M conducted in association
ve,
cti
effe 900
t
cos
800
700
Contributions from: 600
• SP Swedish National Testing

ng
g a
and Research
es h Institute,
titute
tute, Fir
Fire Technology 500
• SINTEF/NBL, Norwegian Fire Research
h Labor
Laboratory
• TNO Building and Construction Research, Centre for Fire Research
• Promat International NV, Belgium Published by:
• Gerco Beveiligingen BV, The Netherlands
2
Content
1 Introduction 3 5 Measurements and interpretation 10
2 Objectives 4 5.1 Overview of fire development 10
2.1 European Commission 4 5.2 Gas temperatures 10
2.2 Technical objectives Runehamar tests 4 5.3 Temperatures in cargo and fire spread 11
3 Preparation 5 5.4 Thermal load on wall at 1 meter above road level 11
3.1 Laboratory tests to predict Heat Release Rate 5 5.5 Heat Release Rate (HRR) 12
3.2 Thermal protection boards 5 5.6 Radiation levels near the fire 13
3.2.1 Boards 5 5.7 Near fire radiation levels and risk of fire spread 14
3.2.2 Pre-qualification of the intended system 6 5.8 Back layering 14
3.2.3 Conclusions of the pre-qualification tests 6 5.9 Results and behaviour of the protective lining 15
3.3 Steel structure 7 6 Conclusions and recommendations 17
4 Tests 8 7 Acknowledgement 18
4.1 Test 1 8 8 Reference list 18
4.2 Test 2 8 Annex A PROMATECT® -T Material description 19
4.3 Test 3 9
4.4 Test 4 9
Foreword
As co-ordinator of the UPTUN project, I am pleased to present this publication on important results of tunnel safety research work.
This work was carried out in the framework of a Swedish national and a European research program on tunnel safety.
Comprehensive large-scale fire tests have been conducted in the abandoned Runehamar road tunnel in the Western part of
Norway in September 2003.
The measurements and preliminary analyses are such that the results will definitely contribute to increase tunnel fire safety levels
in Europe. Therefore the initiative for the actual publication of this document was taken by Promat International with the aim to
inform tunnel related, interested parties, by means of this summary document.
The work presented would not have been possible without the effort and financial contribution of the Swedish Road
Administration, the Norwegian Road Administration, the Swedish Rail Administration, the Swedish Rescue Services Agency,
the Swedish Fire Research Board, the Directorate General Research of the European Commission, the UPTUN partners
(especially SP, NBL and TNO) and the industrial partners, especially the partners who supplied and installed the passive fire
protection system: Promat International N.V., GERCO Beveiligingen B.V.
This document is based on the presentations and papers of the Boras conference on Catastrophic Tunnel Fires, in November 2003.
I would like to thank René van den Bosch (Promat) and Jan Brekelmans (TNO Building and Construction Research) for their
editorial work.
Kees Both, PhD

Coordinator UPTUN project
large scale tunnel fire test Runehamar - September 2003

3
1 INTRODUCTION
Fires in European tunnels in recent years have clearly shown the UPTUN
risks and consequences of fires in large vehicles. Over 20 semi- The UPTUN project concerns the improvements with regard to
trailers, for example, were destroyed in a single fire in the Mont existing tunnels. UPTUN is an abbreviation for ‘cost effective,
Blanc tunnel in 1999. Over 50 people died in these recent fires sustainable and innovative UPgrading methods for fire safety in
in road tunnels. Nevertheless, knowledge of the growth and existing TUNnels’.
spread of fires in semi-trailers is very limited. The most recent The UPTUN project was initiated in September 2002 and is co-
fires in the Eurotunnel (1995), the Mont Blanc tunnel (1999), ordinated by TNO (Netherlands). It is a four year research and
the Tauern tunnel (1999) and the St. Gotthard tunnel (2001) development project.
showed that such fires can develop very high energy releases
(150-600 MW), involving a dozen or so vehicles. The 41 partners of UPTUN originate from 17 European
countries. Several disciplines and professions are incorporated as
Besides the destruction of the tunnel construction and trailers owners, consultants, universities, research organisations and
involved in recent tunnel fires, the tunnel tubes themselves manufacturers. The partners from Eastern countries take part of
were severely damaged by the intensive heat. Due to this, the work for some 10%.
tunnels have been out of service for months and even years It is important to look at tunnels as a system in an
after a fire, causing economic loss for the (surrounding) area. environment. Measures to improve fire safety will therefore be
There is still a huge gap between the outcome of real fires and studied as a system rather than sub-optimised. Positive as well
of small scale tests. There is a need for more detailed knowledge as adverse interaction should be identified. Socio-economic
on how and why various semi-trailer cargos burn so strongly aspects on the wider region have to be taken into account.
and why they spread so quickly. The high heat exposure from The UPTUN project will play a pivotal role in linking up with:
the semi-trailers to the tunnel linings also needs more focus. - various national and international investigations, such as the
The only reasonable way of finding an answer to these EC funded research projects and networks: FIT, Darts and
questions is to carry out systematic large scale experiments that SafeT (see Figure 1)
can provide a better basis for the design of technical systems in - important tunnel associations, such as: International
road tunnels. Tunnelling Association (ITA), World Road Association PIARC
and the United Nations Economic Commission for Europe
The accidents that have occurred in recent years have also and last but not least,
revealed the problems facing the fire and rescue services: they - national projects, such as the Runehamar tests.
have not been able to reach the fire due to the enormous
amount of heat and the dense smoke. The discussions after Figure 1 European research projects and networks "fit" well together.
these accidents have included consideration of equipping fire
and rescue services with mobile fans that can drive the smoke New tunnels
(design)
in a particular direction in order to assist their work. However, 2001-2004
Current
this in turn requires improved knowledge of the effects of such Knowledge
2001-2005 Safety measures
fans. What is the effect on the fire of increasing the Existing tunnels
2002-2006
ventilation? What is the effect on the spread of fire between Guidelines
legislation
vehicles? The situation for the fire services was considered in 2003-2006
these tests, especially the effects of radiation in the vicinity of “linked projects”
the fire, on their ability to approach and fight the fire. National &
International
In the frame of a Swedish national and a European research
program on tunnel safety, comprehensive large scale fire tests
have been carried out in the abandoned Runehamar road
tunnel in the Western part of Norway in September 2003.
Semi-trailer fires, similar to the size of the recent fires in Mont

Blanc Tunnel (France/Italy) and St Gotthard Tunnel
(Switzerland), have been particularly considered. The tests have
been conducted by the Swedish National Testing and Research
Institute (SP) in collaboration with the UPTUN partners: TNO
Building and Construction Research from the Netherlands and See further: [7] (Paper7), [8] (Paper8),
the Norwegian Fire Research Laboratory (SINTEF/NBL). [10] (Ref.1)

4
2 OBJECTIVES
The objectives of carrying out large scale fire tests in the Consequently, a scientifically performed study of semi-trailer
Runehamar tunnel can be described from a political and cargo fires, including systematic variation in the commodity
technical point of view. types, commodity configurations and ventilation conditions as
well as the risk for fire spread between these vehicles would
2.1 European Commission provide information of great importance that is presently
From a political, social and economical point of view tunnels lacking to tunnel authorities, tunnel designers and fire services.
in the Trans-European transport network are very important.
In a political sense mobility is of utmost importance for a
competitive, open European market. This can only be fulfilled See further: General, [12] (Ref.2), [13]
if we can rely on the sustainable growth of a durable and (Ref.3)
reliable transport system. For the end-users it is important
that these tunnels are safe. In case the transport network is
obstructed, it will have an enormous economic and socio-
economic impact. Apart from the direct costs associated with By conducting these full-scale tests the UPTUN partners wanted
reconstruction, wider regions could be out of business for to obtain additionally detailed information about:
extended periods of time. - the influence of ventilation on the peak Heat Release Rate
and fire growth rate,
Due to a growing population and mobility, European - the production of smoke and toxic gasses from various goods
transport networks are extending and more often run and
through various road and railway tunnels than previously. - the possibility for rescue services to fight heavy good vehicle
Recent fires in traffic tunnels, such as Mont-Blanc, Tauern, (HGV) fires.
Gotthard and Channel tunnel, obstructed the open European
market and growth. People lost faith in a safe Trans-European Part of the results of these tests can be found in this document.
road and rail network. These fires in road and rail tunnels Detailed information is available on the attached CD-rom
caused serious loss of life and significant structural damage or in the full Proceedings of the International Symposium on
with serious socio-economic impacts on the wider regional Catastrophic Tunnel Fires, see
economy. http://www.sp.se/fire/Eng/default.htm.
Beside the fires in traffic tunnels, fires in public transport Regarding toxic gases and possibility for
tunnels and underground areas also endangered the faith in rescue, reference is made to [3] (paper 3)
tunnel safety, such as the fires in the funicular tunnel in
Kaprun and in the King’s Cross metro station in London.
To avoid these incidents in the future and to improve tunnel
safety, the relevant Directorate Generals from the European
Commission took the initiatives to draft legislation and to Regarding other activities of UPTUN Work
start up EC funded research projects and networks. Package 2 (fire suppression systems),
All relevant national and international knowledge has been reference is made to [4] (paper 4)
brought together in one of these networks (FIT).
2.2 Technical objectives Runehamar tests

From a technical point of view the project aims to obtain new
knowledge about fire development and fire spread in semi-
trailer cargos and the heat exposure to the tunnel linings in the
vicinity of the fire. There is a lack of systematic studies of the
fire behaviour of semi-trailer cargos. Only two large scale fire
tests using semi-trailer fire loads have been performed in a
tunnel. These tests were performed in 1992 in the EUREKA 499
test program performed in Repparfjord in Norway and
sponsored by European partners. A historic overview of large
scale tests, carried out in the past, is given in [1] (Paper 1),
included on the attached CD-rom.

5
3 PREPARATION
3.1 Laboratory tests to predict Heat Release Rate (HRR) 3.2 Thermal protection boards
Pre-tests consisting of free burning tests under a large hood The tunnel had to be protected with high temperature resistant
system (Industry Calorimeter) at SP’s Fire laboratory were materials because of the expected high thermal output. The
performed prior to the large-scale fire tests. These tests were decision was taken to apply PROMATECT ® -T panels, rather than
carried out in order to obtain some preliminary knowledge for instance a spray mortar.
about the fire development and to estimate the peak HRR of
the commodities used in the large-scale test program. The effects of the intended fire loads on the heat release rate
and the time temperature curve were unknown, prior to the
The set-up of the pre-burn tests is shown in Figure 2a. Three Runehamar tests. Therefore TNO advised to fire test the
tests were carried out using two pallet piles of the commodity. intended construction to the Dutch RWS fire curve, exposing
the panel to multiple fires. This RWS fire curve is still seen as
The height of the piles was 1.5 m, which is about half the the most severe hydrocarbon type of fire, due to its rapid
height of the large-scale fire load. Following type of temperature rise in the first 5 minutes, creating a thermal shock
commodities were tested under the hood system: to the tunnel lining and reaching a maximum temperature of
1 wood pallets and plastic pallets (82/18 %) 1350º C.
2 wood pallets and PUR (polyurethane) mattresses (82/18 %)
3 cartons with PS (polystyrene) cups (81/19 %) The challenge for the PROMATECT ® -T panels can be described
as follows:
- The system should be able to withstand 4 fires with
maximum temperatures going up to approximately 1400º C.
- The temperature criterion on the rock structure of the tunnel
was set to be 250º C. This was perceived to be a safe
temperature for the rock material to minimise damage.
- The system was not allowed to show any integrity failures.
This was applicable to the PROMATECT ® -T panels and the
Reference is made to: [1] (paper 1) sub-frame including the fixation materials.
and [12] (Ref.2) - The system should be easy to install to reduce the installation
time required to install the whole system, and to facilitate
replacement of panels in the unlikely event of damaged
panels (mechanical impact). Promat also wanted to have
the possibility to extract fire exposed panels from the
tunnel, to investigate these in the Promat Research and
Technology Centre (PRTC).
3.2.1 Boards
To enable more than one test, the boards were constructed using
two thinner boards (20 and 25 mm), glued together with inter-
mediate reinforcement (see further, Figure 5). This is not a
standard practical solution but in this case it was chosen to
guarantee the integrity of the panel over more than one extreme
fire test.
Figure 2 Pre-tests SP.

A Free-burning pre-tests under Industry Calorimeter with wood pallets and
plastic pallets.
B Measured HRR.

6
3.2.2 Pre-qualification of the intended system

Figure 3 Pre-qualification tests Promat and GERCO.
Three consecutive fire tests were conducted at the GERCO
laboratory in order to judge:
1 the integrity of the panels after 3 fire tests
2 the integrity of the chosen, easy to install, fixation materials
and system.
As shown on the above Figure 3a, the panels were hung on

threaded rods (M6), which was also the case in the tested
constructions.
Normally such a fixation method is not to be recommended for

tunnels, which are in operation.
The heat sink effect through the steel fixation materials
(threaded rods) was also investigated. Two out of four anchors
A were left unprotected (Figure 3b) and the temperature rise on
the non-exposed face was measured on the protected, as well
as the unprotected threaded rod. As can be seen from Figure 4,
the temperature difference was perceived to be negligible. The
maximum temperatures were 193º C (unprotected) and 174º C
(protected) respectively.
In the tests, the maximum furnace temperature was recorded to

be 1350º C, which is equal to the maximum temperature of the
RWS fire curve. Figure 4 shows the temperature recordings on
the non-exposed face during test 3.
Even after three fire tests the non-exposed face of the boards
B remained well below 200º C, which should be compared to our
maximum allowed temperature on the rock surface of 250º C.
A Construction set-up during fire test.
B Heated PROMATECT ® -T panel and fixations. The heat dissipation in the gap between the PROMATECT ® -T
boards and the rock should also create some additional cooling
effect, leading to the conclusion that the proposed system
should correspond to the established design conditions.
3.2.3 Conclusions of the pre-qualification tests

From the pre-qualification tests the following conclusions were
drawn:
Figure 4 Thermocouple recordings during test 3 on the non-exposed face.
- During three consecutive fire tests no integrity failures were
recorded for the PROMATECT ® -T panels and the sub-frame,
including the fixation materials.
- The maximum temperature on the non-exposed face of the
panels was 186º C, which is well below the criterion of 250º C.
- No major negative influence of penetrations of the threaded
rods were found
Based on the above experiences Promat was confident to
proceed with the described system and offer it to the
Runehamar consortium for use in the full-scale fire test
program in the Runehamar tunnel in Norway. As we know now,
the described system behaved very well, and all partners were
satisfied with the results.

7
3.3 Steel structure

The tunnel width varies from 8.17 m to 9.40 m with a lowest
height of 6.39 m, in the region identified for the fire location.
Figure 5
To determine a best fit location and geometry for the test
Lay-out of steel structure with PROMATECT ® -T boards
set-up, 2400 positions were measured. For ease of installation
a light steel structure was chosen as the support framework
of the PROMATECT ® -T boards.
Within 10 weeks, a 16 ton steel structure with 30 tons of

PROMATECT ® -T boards was produced and built inside the
tunnel by GERCO Beveiligingen B.V. from the Netherlands.
The boards were installed with 4 hooks on steel pipes,

positioned with their long axis in the longitudinal direction of
the tunnel, bearing on the bottom flanges of the truss girders.
These kinds of trusses are traditionally used in greenhouses.
Originally it was planned to drill anchors in the ceiling of the
rock tunnel to install some of these girders. However, due to
some doubts about this connection, it was decided to make a
free-standing structure based on only portal frames. The length
of the thermal isolation is 75 m. Over a part of the walls at
both ends of the structure Promat ceramic blankets were used.
The starting points for the design of the structure were based
on the following assumptions:
- A maximum of 250º C on the non-exposed side of the
thermal board,
- A maximum of 400º C for the protected steel structure and
- A maximum of 600º C the unprotected parts of the steel
structure where the ceramic blankets were installed.
PROMATECT® -T Fixings
Reinforcement
PROMATECT® -T Gerco Glue layer
See further: Video1, Installation, [7]

(Paper7), Brekelmans II, [9] Paper9

8
4 TESTS
In total four tests were performed with a fire in a semi-trailer The commodities are described in more details in Table 6.
set-up. In three tests mixtures of different chosen cellulose and The reason for using furniture is that in the past a test was
plastic materials were used, and in one test a “real” commodity carried out (EUREKA 499) with similar materials and a very high
consisting of furniture and fixtures was used. In all tests the ventilation rate of 6 m/s at the start of the test. This particular
mass ratio was approximately 80% cellulose and 20% plastic. test provides a good point of comparison between the data
A polyester tarpaulin covered the cargo. from the Runehamar tests and the EUREKA tests.
Test nr. Description of the fire load Target Total weight Theoretical Mass ratio
(kg) calorific of plastic
energy (GJ)
1 360 wood pallets measuring 1200 x 800 x 150 mm, 32 wood pallets 10911 240 18%
20 wood pallets measuring 1200 x 1000 x 150 mm and 6 PE pallets
74 PE plastic pallets measuring 1200 x 800 x 150 mm
2 216 wood pallets and 240 PUR mattresses 20 wood pallets

measuring 1200 x 800 x 150 mm and 20 PUR 6853 129 18%
mattresses
3 Furniture and fixtures (tightly packed plastic and Upholstered 8500 152 18%
wood cabinet doors, upholstered PUR arm rest, sofa and (tyres not
upholstered sofas, stuffed animals, potted plant arm rest included)
(plastic), toy house of wood, plastic toys).
10 large rubber tyres (800 kg)
4 600 corrugated paper cartons with interiors 4 wood pallets 3120 67 19%
(600 mm x 400 mm x 500 mm; L x W x H) and and 40 cartons
15 % of total mass of unexpanded polystyrene (PS) with PS cups
cups (18000 cups) and 40 wood pallets (1800 cups)
(1200 x 1000 x 150 mm)
Table 6 Commodities used as fuel in the four tests.
The commodities were placed on particle boards on a storage semi-trailer measuring 10450 mm by 2900 mm. The total height
rack system (see Figure 7, Figure 8 and Figure 9) to simulate a was 4500 mm. The height of the platform floor was 1100 mm.
4.1 Test 1 4.2 Test 2

50
50
10150
1 0150
10450
104 50
Figure 7 Commodity set-up for test 1 (wood pallets and plastic pallets). Figure 8 Commodity set-up for test 2 (wood pallets and PUR mattresses).

9
4.3 Test 3
Figure 9 Commodity set-up for test 3.
4.4 Test 4
The fire was located 560 m from the west entrance and the
wind direction in the tunnel was from east to west.
The cross-section of the tunnel at the site of the fire is shown
in Figure 11. Two small ignition sources, consisting of fibreboard
cubes soaked with heptane, were placed within the lowest
wood pallets (adjacent to the flue between the two pallets) on
the upstream end of the semi-trailer set-up. The tarpaulin was
lifted away during the ignition process. Directly after the
commodity was ignited the tarpaulin was replaced.
At a distance of 15 m from the downstream side of the test
commodity there was a target consisting of the first row of
the same test commodity used in actual test.
Figure 10 Commodity set-up for test 4 (plastic cups in cardboard boxes

on wood pallets).
Figure 11 Cross section of the tunnel at the trailer set-up. See further: Video2, Video3, Video4,
Video5, Fire test, [2] (Paper 2)

10
5 MEASUREMENTS AND INTERPRETATION

5.1 Overview of fire development test 1 test 1
Figure 12a and Figure 12b show the fire development for each
test after respectively 5 minutes and 30 minutes. In test 1 and
test 2 the camera and cargo were on the same position. In test
3 the cargo and camera were moved in the upstream direction
over a distance of respectively 5m and 10m. In test 4 the cargo test 2 test 2
was again moved 5m upstream. According to Figure 12b in all

tests the fire is still burning after 30 minutes, but in particular
in test 1 and test 3 there is still considerable flaming.
test 3 test 3
test 4 test 4
Figure 12 Overview of fire development.

A After 5 minutes
B After 30 minutes A B
5.2 Gas temperatures

Figure 13 Gas temperature.
The four commodities used in the tests were chosen to give
different fire development and maximum heat release rates.
Test 1 with wood pallets and plastic pallets had the highest A
total energy content and gave the highest maximum heat
release rate (see Figure 13a). The large amount of combustible
material also gave a longer period of elevated gas temperatures,
with the highest maximum temperature of 1365º C.
In Figure 13b the gas temperature near the ceiling in test 1
(at + 10 m) is compared to four different standard fire curves.
It can be seen that the increase in gas temperature in the test
with wood pallets and plastic pallets is very rapid and almost
exactly follows the hydrocarbon-curve for about three minutes.
Then the temperature increases even further and more rapidly
than the hydrocarbon-curve and instead follows the RWS curve,
again almost exactly with exceptions for the fast time B
variations and for a period around 20 minutes after ignition
where the measured temperature is higher than the RWS curve.
The RWS curve was developed assuming a tanker fire with
petrol or fuel oil lasting for 120 minutes and giving a heat
release rate of 300 MW. The heat release rate in the tests in the
Runehamar tunnel did not reach 300 MW, but still the
temperature followed the RWS curve very well.
In test 4 only 3120 kg of cardboard boxes and polystyrene cups
were used, potentially creating the lowest calorific energy
output of all tests. However temperatures were recorded to be A Measured gas temperatures close to the fire during the four tests.
in the same magnitude of test 1, although for a shorter period B Gas temperature in test 1 compared with four different standard fire curves.
of time.

11
Thermal protection and spalling 5.3 Temperatures in cargo and fire spread
As shown in Figure 11 the tunnel was protected with Results of the temperature measurements in the cargos are
PROMATECT ® -T boards. This was done for safety reasons to presented in figure 4 in [6] (Paper 6). All tests show
avoid rocks falling from the tunnel structure. A distance of 75 m temperatures between 900º C and 1000º C during 10 to
of the tunnel ceiling was protected, while 25 m of the walls 15 minutes with peak values up to 1200º C in test 1. In test
(near the fire) were protected with these boards. Downstream of 1 the first thermocouple near the fire is heated up about
the board walls, the rock was protected using ceramic curtains 3 minutes after ignition. A mere 7 minutes later the whole
mainly to minimise the flow of hot gases above the protecting cargo is on fire. Test 2 shows an even shorter period of
ceiling. Such hot gases could otherwise affect both the rock 4 minutes between heating up of the first and last
ceiling and the steel structure on which the boards were thermocouple. In all tests the whole cargo is on fire within
hanging. Upstream of the board walls, a distance of 9 m was also 8 to 10 minutes after ignition.
protected with ceramic curtains, although not all the way down
to the road. This was done to keep the back-layering gases below Figure 16 presents ‘the length of the burning part of the cargo’
the protecting ceiling. It was obvious that this protection was as a function of time, based on a temperature of 600º C. Test 1
needed during and after the first test, when large rocks fell down and test 3 show an almost monotonic increase of ‘burning
onto the road both upstream and downstream of the protection length’ with time, indicating a constant fire spread of
(see Figure 14). Downstream of the protection, the tunnel ceiling approximately 18 mm/s for a ‘burning length’ between 1.3 m
was affected almost all the way to the western tunnel entrance. and 6.5 m. This is not the case in test 2 and test 4. Test 4
suggests an even faster fire spread over the same length.
Further analysis of Figure 16 is difficult, because parts of the
Figure 14 Rocks falling down during test 1.
cargo fell down during the tests. This could for instance be the
A B
Figure 16 The length of the burning part of the cargo as function of time.
A upstream of the protected area. B downstream of the protected area.
Figure 15 Temperature upstream of the fire.
cause for the unrealistic behaviour in test 3 where a ‘burning

length’ of 6.5 m seems to appear earlier than a ‘burning
length of 4,5 m.
The rocks falling down upstream of the protection was a result 5.4 Thermal load on wall at 1 meter above road level
of the back-layering taking place in spite of the ventilation. In order to estimate the thermal load on the tunnel wall the
This back-layering was caused by the fact that the velocity heat flux is converted to the temperature of a black body
decreased when the fire intensity increased, increasing the radiating with the same flux as received by the wall. This so
pressure drop over the fire field. The results can be seen in called radiation temperature can be compared with nominal
Figure 15 where the temperatures upstream of the fire during temperature curves that are controlled with plate thermo-
test 1 are presented. It can be seen that 40 m upstream, the couples. The radiation temperature determined in this way is
temperature is well above 100º C during a long time period and slightly higher than the temperature that would have been
as far away as 100 m upstream the temperature is close to measured with a plate thermocouple on the same spot.
100º C. For further details and explanation on the back-layering This is caused by the colder surface of the heat flux meter
phenomenon see section 5.8. resulting in increased convective heat transfer to the sensor.

12
The error in the comparison is relatively low for high heat

fluxes and is estimated to be between 20º C and 50º C.
Figure 17 shows the radiation temperatures on the wall and

some well known standard fire curves. The curves are shifted to
the left in order to facilitate comparison with the fire curves.
Figure 17 Radiation temperatures on the wall 1m above
the floor for all tests.
In test 1 an average temperature of 900º C occurred during
30 minutes with peak values of 1100º C. In test 2 peak values of
1000º C occur. In test 2 and test 4 an average temperature of
800º C can be seen during 15 minutes. Test 3 shows a lower
average over the same period, namely 700º C.
In all tests the thermal load on the wall exceeds the standard
ISO-834 curve used for testing of building materials. In test
1 this lasts 30 minutes and in the other tests approximately
15 minutes. Other fire curves seem more appropriate to
represent the thermal load on the wall during these periods,
as e.g. the hydrocarbon Eurocode 1 curve. Presently tunnel walls
are often left unprotected. The test results clearly show the
necessity of a fire protective lining for wall applications.
5.5 Heat Release Rate (HRR)

A number of different instruments were used to determine the
HRR; 5 bi-directional pressure difference probes, 12 thermo-
couples, 3 oxygen (O2) analysers and 2 carbon dioxide (CO2) /
carbon monoxide (CO) analysers. These measurements are not
included in this document. Reference is made to [1] (Paper 1),
included on the attached CD-rom.
In the first two fire tests, test 1 and test 2, a pulsation of the
fire was experienced during a time period when the fire was
over 130 MW. This created a pulsating flow situation at the
measuring station, where the measurements showed that the
maximum velocity was pulsating in the range of 3 to 4 m/s
down to a minimum in the range of 1 to 1.5 m/s. The frequency
of the maximum velocities was about 45 seconds during this Figure 18 The HRR from the four large-scale fire tests with HGV-trailer
period. Since the air mass flow rate is dependent on the fire load.
velocity measurements the HRR measurements also pulsate
during this period. The HRR curves presented in Figure 18 are
the actual HRR (average for test 1 and 2 during the pulsating
period), although a correction has been made for the
transportation time.

13
Test nr Time from ignition Linear fire growth rate Peak HRR Estimated from laboratory tests
to peak HRR (R=linear regression coefficient) (no target – inclusive target)
(min) (MW/min) (MW) (MW)
1 18.5 20.5 (0.997) 203 (average) 186-217

2 14.3 29.0 (0.991) 158 (average) 167-195
3 10.4 17.0 (0.998) 124.9 -
4 7.7 5 – 70 MW: 17.7 (0.996) 70.5 79-95
Table 19 Peak HRR and fire growth rate from the Runehamar tests.
The fire growth rate appears to be relatively linear for all the 5.6 Radiation levels near the fire
tests when the fire becomes larger than 5 MW and less than The high temperatures give rise to high radiation, which is
100 MW except for test 4 which has a peak HRR of 70 MW. important for the fire spread to other vehicles in the tunnel.
Therefore, a linear curve fit for the different tests was used Another important issue regarding the radiation is how close to
between 5 MW and 100 MW for test 1 to test 3 and between the fire the fire fighters can reach before they are stopped by
5 MW and 70 MW for test 4. The linear regression coefficient R the high radiation. Tests performed with fire fighters in
is shown in parentheses in Table 19 and is found to be very protection clothing indicate that there is a limit approximately
high in all cases (>0.99), indicating a highly linear behaviour 5 kW/m2 exposure above which the fire fighters will have
during this period. Table 19 shows that the wood pallets and difficulty to work and also feel pain after about 5 minutes.
mattresses (test 2) yield the fastest fire development (29 MW/
min), followed by the wood pallets and plastic pallets in test 1 The measurements during the large-scale fire tests, presented in
(21 MW/min). Test 3 and test 4 were found to be very similar Figure 20 show that this limit is exceeded in all of the tests at a
(17-18 MW/min). distance of 10 m from the set-up. The fire fighters not only
need to be able to withstand the radiation, they must also be
Figure 20 Radiation upstream of the semi-trailer set-up able to work in the heat.
(Note the difference in scale).
The radiation level 20 m upstream of the fire is an important
A quantity to determine whether or not the fire brigade can
reach the fire with their water jets. Figure 21 shows the
measured heat fluxes at this distance.
It appears that all heat fluxes remain below the critical level of
5 kW/m2. The fire brigade will therefore be able to approach
the burning cargo up to 20 m and attack the fire.
B Figure 21 Radiation flux upstream 20m behind the fire for all tests.
A 10 m upstream of the semi-trailer set-up

B 5 m upstream of the semi-trailer set-up

14
However, 20 m upstream, in the area were the rock was not Figure 22 Velocities upstream the fire in test 3 and test 4 with predicted
protected against the fire, at 80-100º C, spalling rock has been period of back layering.
recorded, resulting in large blocks of rock falling down in the
area where the fire brigade would be expected to attack the
fire. This would endanger the fire fighters and hamper their
ability to do their work.
5.7 Near fire radiation levels and risk of fire spread

The measured heat fluxes near the fire for all tests together
with the critical level for fire spread of 12.5 kW/m2 are
presented in figure 7 in [6] (Paper 6). In test 1 heat fluxes
on the floor of 250 kW/m2 occur during 15 minutes. In the
same test peak values of 200 kW/m2 and average values of
about 120 kW/m2 on the wall can be observed. At a distance of
5 meter behind the fire the heat flux is still 50 kW/m2.
In all tests the critical level for fire spread is exceeded on the
location 5 m behind the fire. The risk of fire spread to a vehicle
on that location exists therefore in all tests, but for different
lengths of time. In test 1 the risk exists for 55 minutes. In the
other, less severe tests shorter durations of about 7 to 10 minutes
occur. More accurate estimations of the risk of fire spread in
case of a heavy good vehicle fire will be made in the near
future, using more sophisticated radiation models.
5.8 Back-layering
The back-layering of heat and smoke can cause several problems.
It can decrease the visibility both for the people inside the tunnel
and for the rescue personnel. The gases are toxic for people
without proper breathing equipment. The hot gases radiate,
which can affect both the people escaping from the fire and the
fire fighters trying to reach the scene of the fire. As discussed
above, the hot back-layering gases can make rocks fall down and
possibly make concrete start spalling. This can pose a serious
safety problem for the people inside the tunnel.
Temperatures have been measured with thermocouple trees on
Upstream velocities and temperatures were measured in order to 3 upstream locations in the tunnel. In test 1 these trees were
correlate the occurrence of back-layering with the ventilation placed in both lanes and in the middle of the tunnel 100 m
velocity. The velocities were measured with hot sphere upstream of the centre of the fire. In the other tests the trees
anemometers located 150 m upstream, 2.5 m above the floor in were positioned 25m, 50m and 75m upstream in the middle of
both lanes. In addition a bi-directional probe was placed in the the tunnel. Each tree consisted of 5 type K thermocouples
middle of the tunnel, 50 m upstream at a height of 3 m. located 1m, 2m, 3m, 4m and 5m above the road surface.
According to Atkinson [14] the critical velocity to prevent
back-layering should be 2.2 m/s for wide tunnels and 2.5 m/s Figure 13 in [6] (Paper 6) presents the upstream temperatures
for small tunnels for a fire with a heat release rate greater for test 3 and test 4 together with the same predicted period of
than 10MW. back-layering indicated in Figure 21. From these figures it can
In Figure 22 the velocities measured in test 3 and test 4 are be concluded that there is a good correlation between the
shown together with the predicted period of back layering measured and predicted occurrence of back-layering. In all tests
according to Atkinson. a velocity of 2.5 m/s is sufficient to prevent back layering.

15
5.9 Results and behaviour of the protective lining The locations for the different thicknesses of the boards are
Due to irregular shape of the Runehamar tunnel, the indicated in Figure 25. These temperatures are far below the
PROMATECT ® -T boards were installed suspended on a metal design-limit of 250º C, agreed by the partners of this project.
frame; there is no direct contact between the board and the This confirms the results from the laboratory testing of the
rock (Figure 23). The objective of these tests was to determine proposed system (see 3.2.2).
the performance of a range of fire protection in different fire
conditions when no concrete structures were present (concrete
contributes generally to fire resistance of products by a cooling Figure 25 Stability and locations PROMATECT ® -T boards.
effect), no joint protection exists and when the boards were
A
submitted to multiple exposures in successive fires.
Figure 23 Designed fire protection system for the Runehamar tunnel

fire tests.
Note: The metal frame on which the PROMATECT ®-T boards were fixed,
is shown in red.
B
2xPt-T 15 mm board
Pt-T 25 mm sample A & B 10m to fire
Figure 24 Temperatures in the tunnel and on the boards, measured

during the most severe test 1.
Pt-T 25 mm board
sample C ± 0m to fire
A PROMATECT ® -T stability after the test 1.

B Locations of board sample collection.
Some boards were collected (Figure 25b shows the locations)

for detailed investigations on the matrix behaviour after
The Figure 24 illustrates the temperatures developed in the exposure to most intensive fire (test 1) or to successive fires
tunnel during the first fire test (the red curve, temperatures (total of 4). Samples A and B are from overlapped 15 mm
indicated on right side) and the temperatures measured on the boards situated above the vertical side of the fire load, 10m
cold side of the boards (temperature values indicated on left downstream, where the temperature has reached 1365º C.
side of the graph). Sample A was on fire side, sample B on the protected side.
These samples were exposed only to the fire test 1 and will be
These results demonstrate that the fire temperatures reach discussed further.
1365º C in a normal cargo truck fire, no liquid hydrocarbon fuel
being used. The curve overlaps the hydrocarbon curve (first few Scanning Electron Microscopy (SEM/EDX) and X-ray diffraction
minutes) and RWS fire curve (up to 30 minutes) very well. With (XRD) were used to investigate the matrix details and to
this extreme tunnel fire, the maximum temperatures registered establish the profile of temperature evolution inside the
on the cold side of the PROMATECT ® -T 1x25 mm boards and product during the fire tests; the matrix integrity was examined
2x15 mm boards were 210º C and 179º C respectively. by SEM on polished sections, using back-scattered conditions.

16
Thanks to mineral engineering technology, PROMATECT ® -T has Figure 26 Sample A.

“mineral tracers” (a kind of on-site thermometers), that can
1 2
provide information about the evolution of the temperature at
any place on the board. Figure 29 illustrates the evolution of
temperature from the hot to the cold side through the
thickness based on mineral phase transformations and modified
engineered crystal morphologies. Some details are shown in
Figure 26, Figure 27 and Figure 28. A temperature profile can be
established with a maximum of 1150º C at 3 mm depth from
the exposed surface, a maximum of 900º C at 7 mm and a
maximum of 200º C at 20 mm from the exposed surface inside
3 4
the boards. Note that the board was exposed up to 1365°C in
this test.
X-ray diffraction analysis from hot to cold side of the boards

demonstrates that on the exposed face, at a thickness of
3 – 4 mm, a layer of a ceramic insulator was formed at
temperatures between 1150º C and 1350º C. No defects are
created into the matrix during this process (see photo 1 in
Figure 26).
Figure 27 Sample A.
As for the other samples, the SEM/EDX and XRD analysis 5 6
demonstrates the perfect stability of the matrix of the boards
that preserves intact the necessary functions for fire protection.
Figure 29 Position of the sample A and B.
7 8
Figure 28 Sample B.
9 10
Note : the numbers on each picture correspond to the photo number on Figure 29.
See further: [1] (Paper1), [2] (Paper2),

[5] (Paper5), [6] (Paper6), [9] (Paper9)

17
6 CONCLUSIONS AND RECOMMENDATIONS

Gas temperatures and radiation: Back-layering and spalling:
Four large scale fire tests were performed simulating fires in the Back-layering of heat and smoke was registered both visually
cargo of semi-trailers inside a tunnel. The cargo was simulated and using temperature measurements. The observed velocity at
using different mixtures of cellulose and plastics (about 80/20 which back-layering occurs is in good agreement with the
mass ratio). This represents ordinary cargos transported daily on values predicted by Atkinson [14]. Above 2.5 m/s no back-
the roads and thereby also often passing through tunnels. The layering was observed. The back-layering caused rocks to fall
type and amount of combustible materials varied between the down upstream of the passive fire protection (ceiling). This can
tests, but all four combustible mixtures showed very fast pose a risk to both the people trying to evacuate the tunnel
increase in temperature after an initial delay. The results also and for the fire fighting and rescue personnel. It also shows the
show that the tunnel construction and protections need to importance of a suitable protection of the tunnel ceiling and
withstand very high temperature. The standard fire curve best other installations inside the tunnel. A similar problem can
representing the test results is the RWS curve. occur with spalling of concrete used inside tunnels when
exposed to high temperatures. Downstream from the ceiling
Heat Release Rate (HRR): protection the road was covered by rocks that fell down from
The heat flux measurements indicate that it can be difficult for the tunnel ceiling.
the fire fighters to come close enough to the fire to be able to
fight the fire. Without back-layering it is probably possible to PROMATECT ® -T Boards:
fight the fire at 20 m distance with water jets. The tests clearly demonstrated that:
The HRR from four large-scale tests in a heavy good vehicles - the product is capable of resisting the high intensity and high
(HGV) -trailer mock-up in a road tunnel with longitudinal temperatures developed in a tunnel fire (maximum 223 MW,
ventilation were measured. Peak HRRs in the range of 71 to 1365º C);
203 MW (average) were measured. The time to obtain the peak - the mineral engineered matrix, a totally new approach in fire
HRR was found to be in the range of 8 to 18.5 minutes from protection materials, improves the cooling and thermal
ignition. In two of the large-scale tests (test 1 and test 2) protection of the tunnel structures and components.
pulsation of the fire and the smoke upstream of the fire were - temperatures below 200°C can easily be kept for long periods
observed during a period when the fire was larger than of time, on the tunnel structures side with the calculated
approximately 130 MW. The fire growth rate in the range of thickness applied;
5 to 100 MW (70 MW in test 4) is linear for all the tests. - the integrity of the boards was demonstrated, even down to
micron scale, after successive fires;
Fire spread and thermal load on the wall at 1 meter - the boards are easy to install and replace when necessary
above road level: after a fire from the exposed part of the tunnel, securing a
In all tests a rapid fire spread occurs: within 5 to 10 minutes the low cost and short time for repairs and re-opening of the
whole cargo is on fire. A first attempt to estimate the fire tunnel.
spread was partly successful for test 1 and test 3. In test 1,
there is a great risk of fire spread to other vehicles at a distance
of 5 m behind (upstream) the burning cargo during a period of
55 minutes. This risk also exists in the other tests, but for a
shorter duration of 7 to 10 minutes. More accurate estimations
of the risk of fire spread in case of a heavy good vehicle fire will
be made in the near future.
A first attempt was made to correlate the heat flux to the wall
with the intensity of the fire, but more sophisticated modelling
is required. In all tests the thermal load on the wall exceeds the
standard ISO-834 temperature curve for building materials for a
duration of 15 to 30 minutes. Other fire curves seem more
appropriate to represent the thermal load on the wall during
these periods, e.g. the hydrocarbon Euro code 1 curve.
It should be noted that all measurements were taken 1 meter
above road level. Presently tunnel walls are often left
unprotected. The test results clearly illustrate the necessity of a
fire protective lining for wall applications.

18
7 ACKNOWLEDGEMENT
The work presented would not have been possible without the the industrial partners and sponsors of the tests (GERCO
effort and financial contribution of the UPTUN partners Beveiligingen B.V., Promat International N.V.).
(especially SP, NBL and TNO), the Swedish Road Administration,
the Norwegian Road Administration, the Swedish Rescue The publication of this report would not have been possible
Services Agency, the Swedish Fire Research Board, The without the financial support of Promat International N.V.,
Directorate General Research of the European Commission and TNO Building and Construction Research and UPTUN.
8 REFERENCE LIST
Papers [1] – [9] presented at the International Seminar on ‘Catastrophic Tunnel Fires (CTF)’,
Borås, Sweden, 20-21 November 2003, see http://www.sp.se/fire/Eng/default.htm for the full Proceedings.:
[1] Ingason, H. and Lönnermark, A., LARGE SCALE FIRE TESTS IN THE RUNEHAMAR TUNNEL – Heat Release Rate (HRR), SP
Swedish National Testing and Research Institute, Sweden. Paper1, Haukur Ingason I, Ingason II, Ingason III
[2] Lönnermark, A. and Ingason, H., LARGE-SCALE FIRE TESTS IN THE RUNEHAMAR TUNNEL - Gas temperature and radiation, SP
Swedish National Testing and Research Institute, Sweden. Paper2, Anders Lönnermark
[3] Brandt, A.W., PRESENTATION OF TEST RESULT FROM FULL SCALE FIRE TESTS AT RUNEHAMAR TUNNEL, SINTEF, Norwegian
Fire Research Laboratory (NBL), Norway. Paper3, Are Brandt
[4] Opstad, K. and Wighus, R., FIRE SUPPRESSION SYSTEMS FOR ROAD TUNNELS (UPTUN), SINTEF, Norwegian Fire Research
Laboratory (NBL), Norway. Paper4
[5] Anton, O. and Wu, X., RUNEHAMAR TUNNEL FIRE TESTS – UPTUN - FIRE PROTECTION, PROMAT International NV, Belgium.
Paper5, Octavian Anton
[6] Lemaire, A., RUNEHAMAR TUNNEL FIRE TESTS – radiation, fire spread and back layering, TNO Building and Construction
Research, Centre for Fire Research, The Netherlands. Paper6, Tony Lemaire
[7] Brekelmans, J.W.P.M. and Goudzwaard, D. van, RUNEHAMAR tests as part of the UPTUN project, TNO Building and
Construction Research, Centre for Fire Research, The Netherlands Paper7, Jan Brekelmans I, Brekelmans II
[8] Nelisse, R.M.L, THE FIRST YEAR’S RESEARCH RESULTS OF THE EUROPEAN PROJECT UPTUN, TNO Building and Construction
Research, Centre for Fire Research, The Netherlands. Paper8, Mirjam Nelisse
[9] Olst, D. van and Bosch, René van den, PROTECTING THE RUNEHAMAR TUNNEL IN NORWAY WITH PROMATECT“-T AGAINST
MULTIPLE FIRES, AS PART OF THE UPTUN RESEARCH PROGRAMME, PROMAT BV, Tunnel fire protection, The Netherlands.
Paper9, Donald van Olst
[10] Brekelmans, J.W.P.M. TUNNEL SAFETY-RELATED RESEARCH PROJECTS AND NETWORKS FUNDED BY THE EUROPEAN
COMMISSION, PIARC World Congress 2003, September, Durban. Ref.1
[11] Persson, H., BASUTRUSTNING FÖR SKUMSLÄCKNING – Försöksresultat och rekommendationer som underlag för
dimensionering och utförande, SP RAPPORT 1990:36, SP Swedish National Testing and Research Institute,
Borås Sweden, 1990 (in Swedish).
[12] Ingason, H. and Lönnermark, A., PROJECT DESCRIPTION AND PLANNING OF LARGE SCALE TESTS IN RUNEHAMAR TUNNEL,
SP Swedish National Testing and Research Institute, Borås Sweden, August 2003. Ref.2
[13] Ingason, H. and Lönnermark, A., LARGE SCALE TESTS IN RUNEHAMAR TUNNEL, SP Swedish National Testing and Research
Institute, Borås Sweden, September 2003. Ref.3
[14] Atkinson, G.T. et al, FIRE TESTS AND CFD MODELING OF CRITICAL VELOCITY BEHAVIOR, ITC seminar Smoke and Critical
velocity in tunnels 1996, London, p13 - p24.

19
ANNEX A: PROMATECT® -T MATERIAL DESCRIPTION

Mineral engineering and products with engineered Figure 31 The performance of matrix engineered product : thickness /
matrix : fire rate compared with known, standard type of products.
The traditional, commonly used, development and manufacturing

technique for a product aimed at fire protection is the
combination of different inorganic raw materials in order to
obtain a non-combustible, fire resistant product with the
required physico-mechanical characteristics.
Promat’s Research and Technology department developed a new

approach to products’ manufacturing. Selected mineral phases
are synthesised by a controlled crystal growth technology. This
mineral engineering applied by specific manufacturing
technologies enables the achievement of the best performance
for a given application. Not only is the crystallo-chemistry
controlled but also the morphology and the crystals assembling The engineered matrix products can easily secure the same
mode (Figure 30), thus creating a product with specially designed interface temperature with a concrete structure, with only 50%
porosity, density, mechanical performance, thermal conductivity, of the thickness of the other products (Figure 31). Although
dimensional stability in diverse humidity and temperature designed as panels, the engineered matrix allows curving on-site
conditions; a product with an engineered matrix. of a board to cover surfaces with a curvature down to 8 meters
diameter (Figure 32). Easy to install by simple, efficient
techniques (Figure 33, Toulon tunnel), the application of the
Figure 30 Controlled crystal growth technology for engineered matrixes. engineered matrix products can be achieved on existing tunnels
without total prohibition of the traffic.
Figure 32 Maximum on site curving capabilities of the

PROMATECT® -T board.
High performance product:

Figure 33 Toulon tunnel protected with the PROMATECT ® -T boards for
A new generation of high performance products - ceiling, escape routes , smoke extraction ducts and fire doors.
PROMATECT ® -T - was launched with numerous advantages for
the protection of concrete structures, construction of escape
routes, fire doors, cable systems and ventilation systems.
Designed to satisfy all needs, including the most severe fire

situation as described by the RWS fire curve, this product is
not only a barrier to fire or a kind of ceramic protection, but
for a certain period of time can provide an intensive cooling
effect, by cooling down in pre-designed steps the environment
near the board. Afterwards, the board becomes an efficient
thermal insulator at fire temperatures up to 1300º C - 1400º C.

longrine 04 90 14 48 48 - (21611 - 04/17)
ITA Secretariat c/o EPFL - Bât. GC - Station 18 - CH - 1015 Lausanne - Tel. : + 41 21 693 23 10 - Fax : + 41 21 693 41 53
Email : secretariat@ita-aites.org - Web : www.ita-aites.org

ITA WG6 - 2017 FINAL Edition

Uploaded by

Document Information

Original Title

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

ITA WG6 - 2017 FINAL Edition

Uploaded by

Copyright:

Available Formats

Structural Fire Protection For Road Tunnels

ITA Working Group 6

N° ISBN: 978-2-9701122–0-4 ITA REPORT n°18 / april 2017

Structural Fire Protection For Road Tunnels

21611-AITES-REPORT-18-WG6.indd 1 02/06/17 18:33

21611-AITES-REPORT-18-WG6.indd 2 02/06/17 18:33

ITA Working Group 6

21611-AITES-REPORT-18-WG6.indd 3 02/06/17 18:33

4 ITA Working Group 6 Maintenance and Repair

21611-AITES-REPORT-18-WG6.indd 4 02/06/17 18:33

5.4 FIRE RESISTANCE LEVELS................................................................................................................27

Structural Fire Protection For Road Tunnels 5

21611-AITES-REPORT-18-WG6.indd 5 02/06/17 18:33

6 ITA Working Group 6 Maintenance and Repair

21611-AITES-REPORT-18-WG6.indd 6 02/06/17 18:33

Structural Fire Protection For Road Tunnels 7

21611-AITES-REPORT-18-WG6.indd 7 02/06/17 18:33

encountered in the Eureka Tests as shown in

Figure 2.4 : Heat Release from Runehamar tunnel test

Figure 2.3 :Time Dependence Temperature Data from

With all the data available measured from

A simplified model was developed by

• Table 2-1 concerns pool-fires, assuming a

5,5 0 33,3 50,0 634*

5,5 0 50,0 75,0 880*

5,5 0 66,7 100,0 1113*

5,5 0 83,3 125,0 1337*

5,5 0 100,0 150,0 1350

5,5 0 133,3 200,0 1350

Table 2 1: Pool fires

8 ITA Working Group 6 Maintenance and Repair

21611-AITES-REPORT-18-WG6.indd 8 02/06/17 18:33

Comments: related to Area

4,5 1 7,2 5,0 206* 1 car: Fire area = 1,6*4,5 m2

4,5 1 9,0 15,0 558* Minivan: 1,8*5 m2

4,5 1 38,3 35,0 1006* Bus: 2,55*15 ("small"/old)

4,5 1 38,3 50,0 1350 Bus: 2,55*15 (new/ski)

4,5 1 38,3 50,0 1350 LGV: 2,55*15

4,5 1 38,3 100,0 1350 HGV: 2,55*15

Table 2.2 : vehicle fires * Higher if flames touch the wall

2.3 Design Criteria Established

After the publication of its 1999 report, the

2.3.1 PIARC Recommendations

A preliminary and basic criterion to be met

Structural Fire Protection For Road Tunnels 9

21611-AITES-REPORT-18-WG6.indd 9 02/06/17 18:33

Proposed Guidelines by PIARC RWS or the HCinc curve to define design

Main Structure Secondary Structures4

Table 2.3 : PIARC Recommendations

10 ITA Working Group 6 Maintenance and Repair

21611-AITES-REPORT-18-WG6.indd 10 02/06/17 18:33

Fire Safety Engineering These comments open the field to fire

Number Tunnel in Tun-nel Exit

ISO ISO (ii) ISO ISO ISO ISO

ISO ISO ISO ISO ISO ISO

Table 2 5 : Design criteria for different circumstances

Structural Fire Protection For Road Tunnels 11

21611-AITES-REPORT-18-WG6.indd 11 02/06/17 18:33

2.5 Time-Temperature curve Ventilation speed leading to an increase in temperature.

12 ITA Working Group 6 Maintenance and Repair

21611-AITES-REPORT-18-WG6.indd 12 02/06/17 18:33

to investigate a specific type of scenario

Recent research has however revealed the

• Table 2-1 concerns pool-fires, assuming a