Tunnel Brochure PDF

July 2008
Tunnel Fire Protection

For Tunnel Structures & Services
www.promat-tunnel.com
CONTENTS
1. INTRODUCTION
_____________________________________________________________________ 6
2. WHY PROTECT TUNNELS?
________________________________________________________ 6-15
2.1 IMPLICATIONS OF THE EU DIRECTIVE ON STRUCTURAL REQUIREMENTS FOR ROAD TUNNEL ____________ 8
2.2 IMPLICATIONS OF THE EU DIRECTIVE TO TUNNEL STRUCTURES ________________________________ 10
2.2.1 RISK MANAGEMENT ____________________________________________________________ 10
2.2.2 RISK ANALYSIS ________________________________________________________________ 11
2.2.3 COMPUTATIONAL FLUID DYNAMICS (CFD) AND FINITE ELEMENT ANALYSIS (FEA) _______________ 12
2.2.4 ECONOMIC AND ENVIRONMENTAL IMPACT____________________________________________ 14
2.2.5 RISK REDUCTION MEASURES _____________________________________________________ 14
3. TYPES OF FIRE EXPOSURE _______________________________________________________ 24-31

3.1 CELLULOSIC CURVE _________________________________________________________________ 25
3.2 HYDROCARBON CURVE ___________________________________________________________ 25-26
3.3 RABT ZTV CURVE ___________________________________________________________________ 26
3.4 RWS (RIJKSWATERSTAAT) CURVE____________________________________________________ 26-28
FEATURING TESTS & GAS TEMPERATURES AN
PROJECT
3.5 HEAT RELEASE RATE vs TIME-TEMPERATURE CURVE _____________________________________ 29-30
3.5.1 LOCATION OF FIRE IN THE TUNNEL _________________________________________________ 29
3.5.2 VENTILATION SPEED ____________________________________________________________ 29
3.5.3 GRADIENT ____________________________________________________________________ 30
3.5.4 CROSS SECTIONAL AREA_________________________________________________________ 30
3.5.5 TIME/DURATION OF THE FIRE______________________________________________________ 30
3.6 FIRE PROTECTION REQUIREMENTS OVERVIEW __________________________________________ 30-31
4. SPALLING CONCRETE
___________________________________________________________ 32-36
4.1 HEATING RATE AND INTERNAL STRESSES ________________________________________________ 32

4.2 SPALLING OF CONCRETE __________________________________________________________ 32-34
4.2.1 VIOLENT SPALLING _____________________________________________________________ 33
4.2.2 PROGRESSIVE GRADUAL SPALLING (SLOUGHING OFF) ___________________________________ 33
4.2.3 CORNER SPALLING _____________________________________________________________ 33
4.2.4 EXPLOSIVE SPALLING ___________________________________________________________ 33
4.2.5 POST-COOLING SPALLING ________________________________________________________ 33
4.3 TESTING OF SPALLING BEHAVIOUR______________________________________________________ 34
4.3.1 GEOMETRY, PRESTRESSING, CONCRETE MIX AND MOISTURE LEVEL ________________________ 34
4.3.2 AGE OF THE SPECIMEN __________________________________________________________ 34
4.3.3 FURNACE TEMPERATURE_________________________________________________________ 34
4.3.4 INSULATION MATERIAL __________________________________________________________ 34
4.4 FIRE RESISTANCE OF CONCRETE ____________________________________________________ 34-35
4.5 CRITERIA FOR THE FIRE RESISTANCE OF CONCRETE ________________________________________ 36
5. CHOOSING FIRE PROTECTION MATERIALS
___________________________________________ 37
5.1 CONSIDERATIONS WHEN APPLYING PROTECTIVE MATERIALS __________________________________ 37

5.2 BOARD MATERIALS _________________________________________________________________ 37
6. TYPES OF TUNNEL
_________________________________________________________________ 38
6.1 BORED TUNNELS USING CONCRETE SEGMENTS____________________________________________ 38

6.2 IMMERSED TUNNEL _________________________________________________________________ 38
6.3 CUT AND COVER (C&C) TUNNELS _______________________________________________________ 38
7. FIXINGS ________________________________________________________________________ 39-43

7.1 LOCALISED CORROSION AND PITTING IN TUNNEL CONSTRUCTION ___________________________ 39-42
7.1.1 LONG SERVICE LIFE OF STAINLESS STEEL GRADES USED FOR ANCHORS_____________________ 39
7.1.2 STAINLESS STEEL THAT RUSTS __________________________________________________ 39
7.1.3 DUST ATTACKS ANCHORS ________________________________________________________ 40
7.1.4 SALT FILM ON MATERIALS_____________________________________________________ 40-41
7.1.5 CORROSION CAUSED BY POLLUTANT GASES _______________________________________ 41-42
7.2 CRACK WIDTH IN CONCRETE __________________________________________________________ 43
7.3 COMPRESSIVE STRENGTH OF CONCRETE_________________________________________________ 43
8. METHODS OF APPLYING BOARD PROTECTION _____________________________________ 44-53

8.1 AS LOST SHUTTERING ____________________________________________________________ 44-47
8.1.1 PROMATECT-H CEILING SYSTEM, LOST FORMWORK METHOD _________________________ 44-46
8.1.2 PROMATECT-T CEILING SYSTEM, LOST FORMWORK METHOD ____________________________ 47
8.2 POST CONSTRUCTION INSTALLATION TO CONCRETE______________________________________ 47-53
8.2.1 PROMATECT-H OR PROMATECT-T FOR POST CLADDING ________________________________ 48
8.2.2 FIXINGS _____________________________________________________________________ 48
8.2.3 STEEL FRAMING ____________________________________________________________ 48-49
8.2.4 INSTALLATION OF PROMATECT-H OR PROMATECT-T BOARDS____________________________ 49
8.2.5 SUSPENDED CEILINGS, PROTECTIVE MEMBRANE____________________________________ 50-51
8.2.6 SUSPENDED CEILINGS, ESCAPE ROUTE ___________________________________________ 51-52
8.2.7 APPLICATION OF CURVED AND FRAMED SECTIONS __________________________________ 53-54
9. FIRE PROTECTION OF ESSENTIAL SERVICES
______________________________________ 54-58
9.1 AIR SUPPLY AND SMOKE EXTRACTION SYSTEMS ________________________________________ 54-55

9.2 SUSPENDED CEILINGS, SMOKE EXTRACTION PLENUMS ___________________________________ 55-56
9.3 CABLE PROTECTION SYSTEMS ______________________________________________________ 56-57
9.4 SAFE HAVENS _____________________________________________________________________ 58
9.5 FIRE DOORS ______________________________________________________________________ 58
10. FAQs _____________________________________________________________________________ 59

11. TUNNEL FIRE RESEARCH _______________________________________________________ 60-61
11.1 FIT_____________________________________________________________________________ 60
11.2 DARTS __________________________________________________________________________ 60
11.3 UPTUN_______________________________________________________________________ 60-61
11.4 SIRTAKI _________________________________________________________________________ 61
APPENDICES ______________________________________________________________________ 62-67

APPENDIX 1: PHYSICAL PROPERTIES ____________________________________________________ 62-63
APPENDIX 2: WORLDWIDE TUNNEL PROJECT REFERENCE ____________________________________ 64-67
TABLES, GRAPHS & DIAGRAMS

TABLE 1: CASUALTIES IN TUNNEL FIRES ________________________________________________________ 7
TABLE 2: TUNNEL FIRE HISTORY __________________________________________________________ 16-23
TABLE 3: CELLULOSIC FIRE CURVE ___________________________________________________________ 25
TABLE 4: POTENTIAL FIRE DURATION _________________________________________________________ 25
TABLE 5: HYDROCARBON FIRE CURVE ________________________________________________________ 26
TABLE 6: HCM FIRE CURVE _________________________________________________________________ 26
TABLE 7: RABT FIRE CURVE ________________________________________________________________ 26
TABLE 8: RWS FIRE CURVE _________________________________________________________________ 26
TABLE 9: COMMODITIES USED AS FUEL IN THE FOUR TESTS ________________________________________ 27
TABLE 10: FIRE LOAD/FIRE CURVES __________________________________________________________ 29
TABLE 11: OVERVIEW OF FIRE PROTECTION REQUIREMENTS ________________________________________ 31
TABLE 12: IMPORTANT RELATIONS BETWEEN MECHANISMS AND SPALLING PHENOMENA _________________ 33
TABLE 13: FACTORS TO CONSIDER WHEN COMPARING PROMATECT-H OR PROMATECT-T BOARDS__________ 35
TO POLYPROPYLENE FIBRES (PPF) FOR FIRE PROTECTION OF CONCRETE STRUCTURES
TABLE 14: EXTRACT FROM GENERAL BUILDING SUPERVISION APPROVAL DIBT Z-30.3-6 _________________ 40
TABLE 15: EXTREME VALUES OF ATMOSPHERIC COMPONENTS IN ST. GOTTHARD TUNNEL _________________ 42
AND IN POLLUTANT CONCENTRATIONS IN TEST SERIES
TABLE 16: PROMATECT-H PHYSICAL PROPERTIES _______________________________________________ 62
TABLE 17: PROMATECT-T PHYSICAL PROPERTIES _______________________________________________ 63
TABLE 18: WORLDWIDE TUNNEL PROJECT REFERENCE _________________________________________ 64-67
FIGURE 1: EXAMPLES OF FIRE DAMAGE ________________________________________________________ 7

FIGURE 2: HEAT RELEASE RATE (MW)/PROBABILITY DENSITY _______________________________________ 11
FIGURE 3: FIRES AND TEMPERATURES LOCATED AT VARIOUS LOCATIONS ACROSS TUNNEL CROSS SECTIONS ___ 12
FIGURE 4: STRUCTURAL BEAM FOR TUNNEL SYSTEM _____________________________________________ 13
FIGURE 5: TIME/TEMPERATURE CURVES _______________________________________________________ 24
FIGURE 6: SET-UP FOR TEST #1 (WOOD AND PLASTIC PALLETS) _____________________________________ 27
FIGURE 7: SET-UP FOR TEST #2 (WOOD PALLETS AND PUR MATTRESSES) _____________________________ 27
FIGURE 8: SET-UP FOR TEST #3 (WOOD FLAT PACK FURNITURE AND PLASTIC TOYS) ______________________ 27
FIGURE 9: SET-UP FOR TEST #4 (PLASTIC CUPS IN CARDBOARD BOXES ON WOOD PALLETS)________________ 28
FIGURE 10: CROSS SECTION OF THE TUNNEL AT THE TEST LOCATION _________________________________ 28
FIGURE 11: OVERVIEW OF FIRE DEVELOPMENT AFTER FIVE MINUTES _________________________________ 28
FIGURE 12A: MEASURED GAS TEMPERATURES CLOSE TO THE FIRE DURING THE FOUR TESTS _______________ 28
FIGURE 12B: GAS TEMPERATURES IN TEST #1 COMPARED WITH FOUR DIFFERENT STANDARD FIRE CURVES ___ 28
FIGURE 13: TEST SET UP FOR A FULL SIZE TUNNEL SEGMENT ______________________________________ 34
FIGURE 14: INFLUENCE OF TEMPERATURE ON CONCRETE __________________________________________ 36
FIGURE 15: SPALLING OF CONCRETE AFTER ACTUAL FIRE __________________________________________ 36
FIGURE 16: REINFORCEMENT TEMPERATURES __________________________________________________ 36
FIGURE 17: ALLOY STRUCTURE AND STRENGTH PROPERTIES OF STAINLESS STEELS _____________________ 39
FIGURE 18: TYPES OF CORROSION FOUND DURING EXPOSURE OF STAINLESS STEEL SPECIMENS ____________ 40
WITH SALT DROPLETS IN HUMID AIR
FIGURE 19: RELATIVE CLASSIFICATION OF THE CORROSION AFTER LONG-TERM EXPOSURE ________________ 41
AS FACTOR OF THE EFFECTIVE SUM OF THE STEEL GRADES
FIGURE 20: TENSION CRACK CORROSION IN 1.4401 ______________________________________________ 41

FIGURE 21: SELECTIVE CORROSION IN 1.4462 __________________________________________________ 41
FIGURE 22: TENSION CRACK CORROSION IN 1.4439 ______________________________________________ 41
FIGURE 23: LOCALISED PITTING CORROSION IN 1.4539 ____________________________________________ 41
FIGURE 24: BEGINNING TENSION CRACK CORROSION IN 1.4539 _____________________________________ 41
FIGURE 25: MODERATE TOUGH CORROSION IN 1.4529 ____________________________________________ 41
FIGURE 26: CLIMATE CYCLES OF THE 4TH TEST SERIES ___________________________________________ 41
FIGURE 27: BEHAVIOUR OF STAINLESS STEELS IN THE POLLUTANT GAS TEST __________________________ 42
FIGURE 28: MAXIMUM DEPTH OF PITTING CORROSION IN m OF THE ROUND MATERIAL SPECIMENS________ 42
IN THE LABORATORY POLLUTANT GAS TEST
FIGURE 29: BOARDS LAID WITH THE SMOOTH FACE DOWN USING LOST FORMWORK _____________________ 44
FIGURE 30: FIXED STRIP OF PLYWOOD SERVES AS A DATUM FOR THE FIRST ROW OF BOARDS ______________ 44
FIGURE 31: BOARD JOINTS ARE SIMPLY ABUTTED, NO SPECIAL TREATMENT REQUIRED____________________ 44
FIGURE 32: BOARDS CUT TO AN ANGLE AT THE HAUNCHES _________________________________________ 44
FIGURE 33: PRE-MARKED SCREW LOCATIONS ON THE BOARDS _____________________________________ 45
FIGURE 34: FIRST LAYER ON REINFORCEMENT INSTALLED PRIOR TO SCREW INSERTION __________________ 45
FIGURE 35: PRINCIPLES OF SCREW INSERTION __________________________________________________ 45
FIGURE 36: SCREW INSERTION ______________________________________________________________ 45
FIGURE 37: SCREW INSTALLED AND PROTECTED BY REINFORCEMENT ________________________________ 45
FIGURE 38: EXAMPLE OF SCREW PATTERN OF LOST FORMWORK ____________________________________ 45
FIGURE 39: THE POURING OF CONCRETE ______________________________________________________ 46
FIGURE 40: EXTRACTED FORMWORK _________________________________________________________ 46
FIGURE 41: OPENING OF MANHOLES _________________________________________________________ 46
FIGURE 42: ANCHOR SYSTEM FOR JET-FANS ___________________________________________________ 46
FIGURE 43: POST INSTALLATION OF SINGLE LAYER PROMATECT BOARDS _____________________________ 47
FIGURE 44: POST CONSTRUCTION METHOD IN ZEEBURG TUNNEL, NETHERLANDS _______________________ 47
FIGURE 45: APPLICATION OF CURVED PROMATECT-T TO CAST IRON TUNNEL LINING SECTIONS _____________ 48
IN CLYDE TUNNEL, GLASGOW, SCOTLAND
FIGURE 46: PANEL LIFTING HOIST ____________________________________________________________ 49
FIGURE 47: DRILLING OF CONCRETE __________________________________________________________ 49
FIGURE 48: SUSPENDED CEILINGS, PROTECTIVE MEMBRANE _______________________________________ 50
FIGURE 49: PRINCIPLES OF ESCAPE ROUTE FOR CIRCULAR TUNNELS _________________________________ 51
FIGURE 50: LOADBEARING CEILING CONSTRUCTION ______________________________________________ 52
FIGURE 51: SECTION THROUGH FLAT CONCRETE SLAB ____________________________________________ 53
FIGURE 52: SECTION THROUGH CURVED CONCRETE SECTIONS ______________________________________ 53
FIGURE 53: TYPICAL TUNNEL SERVICES SYSTEMS _______________________________________________ 54
FIGURE 54: EFFECTS OF AIR FLOW ON CAR FIRE ________________________________________________ 55
FIGURE 55: SMOKE EXTRACTION PLENUMS ____________________________________________________ 55
FIGURE 56: SERVICE TUNNEL _______________________________________________________________ 57
FIGURE 57: PROMAT CABLE PROTECTION SYSTEMS ______________________________________________ 57
FIGURE 58: STEEL FIRE DOORS WITHIN TUNNELS ________________________________________________ 58
1. INTRODUCTION
Fires in tunnels are a major hazard to human life. Tunnel fires also cause costly damage to surrounding
infrastructure. Limited escape facilities and difficulties encountered by intervention forces in gaining
access to the tunnel fire call for extensive safety arrangements which must be complementary and
mutually coordinated.
Tunnels and underground transport facilities are important means of communication, not only in terms
of shorter journey times but increasingly out of consideration for the local population and the
environment. Generally speaking, important underground transport links are expected to be available
without any restrictions and to operate smoothly round the clock. Interruptions due to accidents,
technical malfunctions or maintenance work quickly cause traffic jams and delays, and figure in
transport policy statistics as economic losses.
Rising traffic densities and the growing demand for underground communication links result in a higher
probability of accidents, injuries and damage. Added to this are other factors which increase the
potential hazards of traffic tunnels:
the increasing length of modern tunnels
the transportation of hazardous materials
two-way traffic (with undivided carriageways)
higher fire loads due to growing traffic volumes and higher loading capacities
mechanical defects in motor vehicles
When considering tunnels, it is usually in relation to road and rail infrastructure. However, use of the
word tunnels can be slightly misleading, as the following information can apply equally to pedestrian
walkways, underground rail stations, underground car parks etc...in fact to any concrete structure.
Although this document refers to tunnels throughout, all information also applies to underground spaces
of any description.
It is usually assumed that because a structure is constructed using concrete, that it is inherently fire
resistant, and therefore requires no additional fire protection measures to be taken. Unfortunately,
experience over the years has shown that this is not necessarily the case and consideration must be
given to the performance and behaviour of concrete structures under fire conditions. In addition, where
tunnels and underground spaces are concerned, consideration must also be given to the provision of
services protection, e.g. smoke extraction systems, protection to cables and wiring providing power to
emergency equipment.
This handbook is intended to provide some background into the behaviour of concrete under fire
conditions, to show proven methods of protecting structures against fire, and of providing protection to
services within tunnels and underground spaces.

There are three reasons for providing protection against fire within tunnels. First, there is the matter of
life safety. This is not necessarily a function of structural performance under fire although a collapsing
structure would not enable people to exit a structure in safety but more to do with the function of
services such as emergency lighting, smoke extraction systems and so on.
Within Europe alone, in the past decade or so, there have occurred within road and rail tunnels at least
10 major fires, and countless minor fire situations. These fires have resulted in a major loss of life (221
dead in four fires that took place over a period of just two years) and in all cases significant structural
damage occurred, not to mention substantial economic costs to the community.
As an example, outlined below are some of the tunnel fires which have occurred during the 1990s and
earlier this decade, and the resultant death toll. TABLE 2 on pages 16-23 gives an extensive overview of
tunnel fire history.
TABLE 1: CASUALTIES IN TUNNEL FIRES
Location
Casualties
Location
Casualties
Bosnia
35
Hokuriku Tunnel, Japan
34
Second, there is the performance of the

Mont Blanc, France
39
Pecorile Tunnel
structure itself. Will it remain in-situ? Will it
collapse, possibly causing collateral
Tauern Tunnel, Austria
12
OShimizu Tunnel
damage to other parts of the structure and
Vierzy Tunnel, France
108
Salang Tunnel, Afghanistan
injuries to people passing by? In the MontBlanc-Tunnel fire, there was severe spalling
Pfnder Tunnel, Austria
3
Kings Cross, England
of the structural concrete. During the fire
Huguenot Tunnel
3
Isola delle Femmine, Italy
which occurred inside the St. Gotthard
Tunnel in 2001, a 250m long section of the
Nihonzaka, Japan
7
Velsen, Netherlands
structure actually collapsed, hampering the
St Gottard Tunnel, Switzerland
11
Kaprun, Austria
activities of the rescue services. Although
both these tunnels pass through rock,
localised collapse or spalling although potentially costly and inconvenient, did not endanger persons
located away from the damaged areas. If these tunnels had been of the immersed type, the structural
damage could have resulted in flooding of the tunnels, with all the associated implications.
8
16
700
31
5
5
155
It should be noted that after the fire in the Channel Tunnel, the only thing standing between total loss
and a situation where effective repair could be carried out was the thin grout layer between the concrete
structure and the water bearing rock layer, so severe was the spalling of the concrete. A very slim margin
to rely on, but a risk which could easily have been alleviated had the correct passive fire protection
systems been included, complementing the active systems that were installed.
Thirdly, there is the economic damage caused as a result of the failure of a tunnel. This economic cost
is not related solely to the repair or rebuilding of the structure; more usually it is the knock-on impact of
loss to business and traffic diversions etc. which result in the largest costs.
A prime example is the Channel Tunnel fire. Economic damage was estimated to be over twice the cost
of the actual tunnel repairs. The direct repairs to the tunnel cost an estimated 87 million while the
additional costs in lost business, replacement of infrastructure, materials (e.g. lorries, train carriages etc)
together with the impact of the tunnel closure on other, unrelated businesses brought the economic loss
alone to some 215 million.
FIGURE 1: EXAMPLES OF FIRE DAMAGE
Using Mont Blanc Tunnel as a comparison
for a simple road tunnel, the differences are
not so marked, with the cost of repair being
estimated at approximately 206 million and
the economic cost at some 250 million.
However, the socioeconomic impact has to
be considered on a wider basis rather than
simply the tunnel itself. The estimates of the
effects on the local Italian economy around
the area of the Mont Blanc Tunnel were
estimated at 1.75 billion. Therefore, in any
risk analysis, the socioeconomic costs
need to be accurately identified and
carefully assessed.
Example 1 Moorfleet Tunnel, Germany
Example 2 Mont Blanc Tunnel, Italy
Example 3 Velsen Tunnel, Netherlands
Example 4 Channel Tunnel, UK
In terms of fire protection within tunnel and

underground systems, the following items
should require some consideration.
Enhancing the fire resistance
of the structure
Air supply systems
Smoke extract systems
The provision of fire and smoke
resistant safe havens in long tunnels
Active and passive detection systems
Fire extinguishing systems
Fire doors
Warning and alarm systems

2.1 IMPLICATIONS OF THE EU DIRECTIVE ON STRUCTURAL REQUIREMENTS
FOR ROAD TUNNEL
Following the spate of aforementioned catastrophic fires in European Road Tunnels, it became apparent
that the international tunnelling community had severe reservations on the safety and operations of
tunnels. Therefore in 2001 a paper was released by PIARC (The World Road Association) titled
RECOMMENDATIONS OF THE GROUP OF EXPERTS ON SAFETY OF ROAD TUNNELS

The document presents the following prognosis.
To ensure safety in road traffic, the necessary structural, technical and organisational measures need to
be taken. All safety measures have to correspond to the latest technology and apply to all concerned,
i.e. to road users, traffic control and emergency services, infrastructure and vehicles.
The following objectives have been set for attaining the optimal level of safety in road tunnels:
Primary objective Prevention
To prevent critical events which endanger human life, the environment and tunnel installations.
Secondary objective Reduction of consequences
As a result of events such as accidents and fires; to create the ideal prerequisites for
people involved in the incident to rescue themselves,
the immediate intervention of road users to prevent greater consequences,
ensuring efficient action by emergency services,
protecting the environment and
limiting material damage.
The document concludes with the following statement:
Fires in tunnels not only endanger the lives of road users, they can also cause damage to structural
components, installations and vehicles, with the result that the tunnel concerned may have to be closed
for a considerable length of time.
The above paper made tunnel stakeholders acknowledge the frailty of the safety issues associated with
the operation of their own specific tunnel and in particular the concerns with mitigating the
consequences of structural damage and the impact this has on the environment due to extended
diversion routes.
Notwithstanding the above report and the consequences of the catastrophic fires it was also recognised
within the European tunnelling community that a wide range of operational and safety standards,
regulations and structural requirements existed in many different countries. The community believed that
this led to confusion and had to be standardised.
This standardisation process led to the introduction of the EU Directive in 2004 in a document titled
DIRECTIVE 2004/54/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

ON MINIMUM SAFETY REQUIREMENTS FOR TUNNELS
IN THE TRANS-EUROPEAN ROAD NETWORK
OBJECTIVE OF THE EU DIRECTIVE
The objective of the directive was primarily to harmonise and introduce minimum safety standards
for road tunnels in Europe and was primarily targeted at tunnel stakeholders. The document also
intended to make stakeholders fully aware of the risks to life, structure and the economic and
environmental impacts associated with the operation of unsafe tunnels.
Article 1 of the document states:
The Directive aims at ensuring a minimum level of safety for road users in tunnels in the
Trans-European Road Network by the prevention of critical events that may endanger
human life, the environment and tunnel installations, as well as by the provision of protection in
case of accidents.
The aim of this statement was to raise stakeholder awareness of risk and the consequences of
risk with special significance placed upon the risk of fire and the consequences to life, structure
and the environment.
Article 3 gives guidance on how to reduce the consequences of risk namely by:
Implementation of Risk Reduction Measuresthe efficiency of these measures shall be
demonstrated through a risk analysis in conformity with the provisions of Article 13.
This statement suggests that Risk Reduction Measures need to be implemented but their
Efficiency and Performance Requirements needs to be assessed through Risk Analysis
techniques. However, it is pointless incorporating risk reduction measures unless the effectiveness
of these Risk Reduction Measures is known.
Article 13 states:
Risk Analyses, where necessary, shall be carried out by a body which is functionally
independent from the Tunnel Manager. A risk analysis is an analysis of risks for a given tunnel,
taking into account all design factors and traffic conditions that affect safety, notably traffic
characteristics and type, tunnel length and tunnel geometry, as well as the forecast number of
heavy goods vehicles per day.
In conclusion, the objective of the EU Directive was primarily to make tunnel stakeholders more risk
averse in generic operational risks with particular reference to the consequences of fire.
Continued from page 9
2.2 IMPLICATIONS OF THE EU DIRECTIVE TO TUNNEL STRUCTURES

With reference to tunnel structures, the European Directive states:
The main structure of all tunnels where a local collapse of the structure could have catastrophic
consequences shall ensure a sufficient level of fire resistance.
In this statement specific reference is made to the main structure forming the tunnel and the need for a
sufficient level of fire resistance. The ambiguity here is what can be deemed to be a sufficient level of
resistance. This level of resistance can only be assessed if we know what the magnitude of fire risk.
This fire magnitude can only be assessed through risk analysis and the document refers to Article 13
above. The EU Directive repeatedly refers to Article 13 in an attempt to harmonise standards. In principle
the EU Directive introduces Risk Management techniques to introduce minimum safety standards for
roads on the Trans-European Road Network.
2.2.1 RISK MANAGEMENT

With specific reference to tunnel structures, we know that
THE RISK = FIRE

THE CONSEQUENCES = STRUCTURAL COLLAPSE
But how do we derive the solution to satisfy the requirement of ensuring a sufficient level of
fire resistance?
The EU Directive, as explained above, refers to Risk Management techniques and makes specific
reference to risk analysis but other Risk Management tools are required to derive a solution. Risk
Analysis and in particular, reference to tunnel structural integrity unfortunately only identifies the
probability and magnitude of the risk. It does not conclude the consequence.
Promat International in joint partnership with leading consultants in the field of Risk Management of
structures has developed a suite of tools which allows the tunnel stakeholders to make key
decisions to ensure compliance with the EU Directive. These include:
Risk Analysis risk impact (probability and magnitude)
CFD & FEA Modelling structural impact and consequence
Consequential Analysis economic and environmental impact
Resulting in the optimisation of
Risk Reduction Measures
However, it must be emphasised that not all tunnels require risk reduction measures in order to
meet the requirements of the EU Directive. Indeed if the probability and magnitude of the risk is
small, consequential analysis is not required. This eliminates the need for risk reduction measures
and the Risk Management function is therefore complete. The Risk Analysis tool is therefore a
fundamental and powerful tool in setting the constraints for Risk Management and similarly the
need or otherwise for structural Risk Reduction Measures.
10
2.2.2 RISK ANALYSIS

Almost all risk assessment tools use the explicit risk assessment formula
RISK = FREQUENCY x CONSEQUENCE

However, this approach makes no specific reference to the magnitude of the risk. Furthermore
when considered with specific reference to fires in tunnels, it cannot determine consequence. We
know the potential consequence to tunnel structures from the risk of fire but this does not give
guidance in assessing the level of the risk reduction measure. This can only be achieved in
assessing the frequency (probability) and magnitude of the fire risk (fire load or HRR, the heat
release rate).
Promat International with its partnering consultant IEB Consulting Ltd have developed a
Risk Analysis tool which allows the probability and fire magnitude risk to be determined for any type
of tunnel. This approach, originally pioneered in the UK following the introduction of the EU
Directive, has been used widely throughout the world to derive design fire sizes for structural
resistance in tunnels.
Output from this Qualitative Risk
Analysis (QRA) model results in a
Probability Fire Size Matrix
(see at right).
PROBABILITY
DENSITY
10
HRR (MW)
Probability
Years
0.325
3.08
15
0.056
17.92
25
0.056
17.79
50
0.068
14.60
100
0.023
42.92
100
1000
0.1
0.01
y = 2.4105x-1.5968
0.001
0.0001
HEAT RELEASE RATE (MW)

FIGURE 2: HEAT RELEASE RATE (MW)/PROBABILITY DENSITY
Using a balanced approach, analysis of the matrix determines the design fire size and as shown
in FIGURE 2 sets the Structural Design Criteria in the consequential analysis using Computational
Fluid Dynamics (CFD) & Finite Element Analysis (FEA) tools to assess the need or otherwise for Risk
Reduction Measures.
To summarise, the risk reduction process can be concluded at this stage if the Probability Fire
Size Matrix shows that the risk is small. In the above case it was concluded that the probability of
the 100MW fire lay within the design life of the tunnel. CFD & FEA was then used to assess the
structural performance from such a fire load.
11
2.2.3 COMPUTATIONAL FLUID DYNAMICS (CFD) AND FINITE ELEMENT ANALYSIS (FEA)
Following risk analysis, if it is concluded that the probability and magnitude of the fire risk may result
in consequential structural damage, an assessment of this consequential damage needs to be
undertaken. IEB Consulting Ltd and TNO/Efectis have pioneered the coupling of Computational
Fluid Dynamics (CFD) and Finite Element Analysis (FEA) to assess the structural damage resulting
from the output of the risk analysis.
Coupling of these two design tools allows individual structural components to be assessed in detail
with fires located at various locations across tunnel cross sections.
FIGURE 3: FIRES AND TEMPERATURES LOCATED AT VARIOUS LOCATIONS ACROSS TUNNEL CROSS SECTIONS
rating
Sepa
wall
er
ld
rd
Ha
u
ho
Fire
The CFD component develops the temperature constraints at the boundary of the structure for a
fire at any location in a tunnel important for wide tunnels formed from many structural
components.
Cross section x = 0.000m
1150C
1200C
1250C
1200C
1300C
1200C
1350C
1250C
1350C
1400C
1300C
1350C
1400C
1450C
1150C
1050C
1000C
1100C
1200C
The FEA component allows detailed time dependent failure analysis to be assessed for all
components forming the structure and allows detailed assessment up to the point at which the
structure is no longer self supporting and failure mechanism begins. For complex structures such
as cut & cover tunnels, this is crucial to determine the weakest element.
12
For complex structures such as cut & cover and immersed tube tunnels, specific failure criteria can
be derived for individual structural components.
This CFD & FEA approach has recently been accepted by the European tunnel community as
current best practice in assessing structural resilience of tunnels and was recently presented at a
symposium on catastrophic fires in tunnels.
FIGURE 4: STRUCTURAL BEAM FOR TUNNEL SYSTEM
Real beam
Subdivided beam
Mechanical model, truss elements
2D, mesh for thermal analysis
Mechanical model
WP4 TUNNEL SYSTEM STRUCTURAL RESPONSE

Objectives and results:
Insight into the structural performance of load bearing elements,
Define structural procedures to reduce critical behaviour,
Investigate damage mitigation of the load bearing structures,
Preserve the functional characteristics of structures,
Optimise repair and recovery procedures.
Please contact Promat for more information on this analysis by TNO Building & Construction
Research in collaboration with Uptun.
Following consequential analysis, a full assessment of risk reduction measures can be evaluated. If
the analysis concludes that no significant structural damage is likely to occur, then no risk reduction
measures are required. However, if the analysis concludes that severe structural damage is likely as
a consequence of the fire magnitude derived from the risk analysis, risk reduction measures will then
be required.
Output from this tool allows the performance requirements for each component forming the tunnel,
structure to be determined, resulting in an optimisation of the risk reduction measure. This enables
the designer to provide the most cost-effective solution for benefit of the stakeholder.
This approach, again pioneered following the introduction of the EU Directive has been used
universally throughout Europe to assess the performance of tunnel structures.
13
2.2.4 ECONOMIC AND ENVIRONMENTAL IMPACT

Another factor in assessing the need for structural resilience of tunnels is the need to assess the
economic & environmental impact following failure due to fire.
As described on pages 6 and 7, the loss of tunnel operations due to structural collapse can have
severe economic and environmental impact on local and national communities.
IEB Consulting Ltd, partners with Promat International have developed an Excel model which allows
an economic dis-benefit analysis for any tunnel to be assessed. This analysis determines the
potential loss in operational revenue and combined with the CFD & FEA model, the environmental
and economic impact to the community and country to be evaluated. This could be a major factor
in assessing the need for structural or asset protection against the risk of fire.
2.2.5 RISK REDUCTION MEASURES
In real terms, for existing tunnels, there are two main approaches to introducing Risk Reduction
measures. These are:
THE BOARD SOLUTION
and
THE SPRAY APPLIED SOLUTION
Output from the CFD & FEA analysis will set the design constraints for either solution and may even
be a combination of both types in order to provide the best and cost effective approach for any
specific tunnel. The recent acquisition of Cafco International by Promat assures flexibility in
achieving the best solution.
14
CONCLUSION
Existing EU Legislation & Current Best Practice (PIARC) requires tunnel stakeholders to assess
structural risks in their particular tunnel. Current experience suggests that all stakeholders should
assess these structural risks by the use of risk assessments. However, risk assessments do not fully
assess the complete understanding of the risk and therefore consequential analysis is required.
The suite of Design Tools developed by IEB Consulting Ltd and TNO/Efectis, allows Promat
International to assist stakeholders to appreciate their obligations regarding Legislative Requirements &
Best Practice.
These tools will also allow Promat to work through team integration with stakeholders in assessing the
probability, magnitude, consequence, economic and environmental impact of the fire risk for any type
of tunnel.
These Design Tools include:
Quantified Risk Assessments,
CFD & FEA Analysis,
Economic & Environmental Impact Analysis, and
Design of Risk Reduction Solutions.
They will enable stakeholders to fully assess, through a Risk Management process, compliance with the
EU Directive and engage in current best practice.
15
TABLE 2: TUNNEL FIRE HISTORY

Consequences
Duration
of fire Damaged Structures and
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
Afghanistan
1982
Mazar-e-SharifKabul Salang
2,700m
Soviet Military column,

at least one petrol truck
Gas tanker
explosion
Not
available
Australia
2007
Burnley
Melbourne
3,400m
Car/truck collision
Fire due to
collision
1 hour
Cars and
trucks
Not available
Austria
2002
Tauern
Salzburg
6,400m
Lorry
Faulty engine
Not
available
Not
available
Severe smoke
production
Austria
2001
Gleinalm
A 9 near Graz
8,320m
Coach
Short circuit
>1 hour
1 coach
Severe smoke
production over 3km
Austria
2001
Gleinalm
A 9 near Graz
8,320m
Swedish tourist
coach
Not available
Not
available
Not
available
Not available
Austria
2001
Gleinalm
A 9 near Graz
8,320m
Coach
Not available
Not
available
Not
available
Not available
Austria
2001
Gleinalm
A 9 near Graz
8,320m
Car
Front collision
lorry-car
Not
available
Not
available
Not available
Austria
2001
Tauern
Salzburg
6,400m
Cars
Head on collision
of two cars
Not
available
2 cars
Not available
Austria
2000
Kitzsteinhorn
Kaprun
funicular tunnel
3,300m
Passenger train
Hydraulic oil leak

onto heater
Not
available
1 train
Line closed for

over 1 year
Austria
2000
Tauern
Salzburg
6,400m
HGV
Not available
hour
1 HGV
Not available
Austria
1999
Tauern A10
Salzburg
Spittal
6,400m
Lorry loaded
with paint
Austria
1995
Pfander
6,719m
Lorry with trailer
Collision
Austria
1989
Brenner
200m
Dangerous goods
exploded during
construction
Austria
1987
Tanzenberg
2,400m
Austria
1986
Herzogberg
Austria
1984
Azerbaijan
Not
Severe damage to
available
structure
Front-rear collision
16 lorries,
15 hours
4 cars and 2 lorries
24 cars
Serious damage
1 hour
1 lorry, 1
van, 1 car
Serious damage
Dangerous goods
7 hours
Not
available
Not available
Car
Suicidal car driver
Not
available
1 car
Significant damage to
tunnel structure and
dense smoke
2,000m
HGV
Overheated brakes
Not
available
1 HGV
1 HGV
Felbertauern
5,000m
Coach
Overheated brakes
>1 hour
1 coach
Damage to tunnel
lining >100m
1995
Baku
Not
available
Railway/metro train
Electrical fault
at rail car
Not
available
1 train
Severe smoke
production
Belgium
2004
Kinkempois
600m
HGV
Not available
Not
available
1 HGV
Closed for
several days
Belgium
1987
Brussels
Underground
Not
available
Station fire
Not available
Not
available
Not
available
Dense smoke
Canada
2000
Montreal
Metro
Not
available
Cable fire
Cable
6 hours
Not
available
Electrical system,
severe smoke
production
Canada
2000
Toronto
Metro
Not
available
Railway/metro train
Not available
Not
available
1 train
Line closed
for 24 hours
Canada
1997
Toronto
Metro
Not
available
Train
Not
Rubber matting under
the track caught fire available
Rubber
matting
Severe smoke
production
16
Consequences
Duration
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
Canada
1976
Christie Street
Metro Montreal
Not
available
Station fire
Arson attack
Not
available
Not
available
Damage
>$3 million
Canada
1974
Rosemont
Metro Montreal
Not
available
Train fire
Short circuit
Not
available
1 train
Not available
Canada
1971
Henri Bourassa
Montreal
Not
available
Train fire
Collision with end

of the tunnel
Not
available
1 train
Not available
China
1998
Gueizhou
Guiyang/
Chansha
800m
Train
Exploding gas
canisters
Not
available
1 train
Tunnel collapsed
Denmark
1994
Great Belt
Korsor during
construction
Not
available
Tunnel boring
machine
Not
Explosion at TBM due
to leaking hydraulic oil available
1 TBM
Severe damage
to tunnel lining
France
2004
Dullin
Chambery
1,500m
Coach
Engine
compartment
1 hour
1 coach
Not available
France
2003
Cret dEau
4,000m
Train
Sleeper carriage
Not
available
1 train
Not available
France
2003
Mornay
2,600m
Train
Passenger
carriage
5 hours
1 train
multiple
cars
Not available
France
2002
A86 Versailles
Under
construction
Not
available
Cargo train
Engine exploded
6 hours
1 cargo
train
Not available
France
2000
Toulon
1,850m
Construction vehicle
Collision of two
construction vehicles
17
4 hours construction
Not available
vehicles
France
1994
Castellar
570m
HGV carrying
waste paper
Tyre caught fire
Not
available
1 HGV
Not available
France
1986
LArme
Nice
1,105m
Lorry with trailer
Braking after
high speed
Not
available
1 lorry,
4 cars
Equipment
destroyed
France
1985
Paris Metro
Not
available
Station fire
Rubbish fire
Not
available
Not
available
Not available
France
1983
Frejus
12,900m
HGV loaded with

plastics
Gearbox fire
2 hours
1 HGV
Severe damage
to tunnel lining
over 200m
France
1979
Paris Metro
Not
available
Train fire
Short circuit
Not
available
1 train
Heavy smoke
France
1977
Paris Metro
Not
available
Station fire
Not available
Not
available
Not
available
Not available
France
1976
Crossing
BP-A6 Paris
430m
Lorry with drums of

16 tons polyester film
High Speed
4 hours
1 lorry
Serious damage
over 150m
France
1976
Porte
dItalie B6
430m
HGV carrying 16 tons

polyester plastic
Engine fire
1 hour
France
1975
Chteau de
Vincennes
Metro
Not
available
Train
Short circuit
under car fire
Not
available
France
1973
Porte dItalie
Metro
430m
Railway carriage
Arson attack
Not
1 carriage
available
Not available
France
1972
Vierzy
Not
available
Passenger train
Tunnel collapse
Not
available
1 train
Not available
France
1971
Crozet
Not
available
Goods train and

hydrcarbon fuel train
Collision and
derailment
Not
available
2 trains
Not available
Not
Tunnel lining
available destroyed over 150m
1 train
Not available
TABLE 2: TUNNEL FIRE HISTORY Continued from page 17

Consequences
Duration
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
France
1971
Paris Metro
Not
available
Train fire
Arson attack
Not
available
1 train
Not available
France
1921
Batignolles
1,000m
Passenger train
Collision with
stationary train
Not
available
2 trains
Not available
France
1903
Couronnes
Metro Paris
Not
available
Train fire
Electrical fault
Not
available
2 trains
Not available
France
1842
Mendon
Not
available
Train fire
Not available
Not
available
Not
available
Not available
France/Italy
2004
Frejus
12,900m
HGV
1 HGV
Not available
France/Italy
1999
Mont Blanc
11,600m
Lorry with 12 tons flour

and 9 tons margarine
Oil leakage motor,

overheating
>53
hours
23 lorries, 10
cars, 1 motorcycle etc
Serious damage,
tunnel reopens
France/Italy
1990
Mont Blanc
11,600m
Lorry with
20 tons cotton
Motor
Not
available
1 lorry
Equipment
destroyed
France/Italy
1988
Mont Blanc
11,600m
HGV
Not available
Not
available
1 HGV
Not available
France/Italy
1981
Mont Blanc
11,600m
HGV
Engine fire
Not
available
1 HGV
Not available
France/Italy
1978
Mont Blanc
11,600m
HGV
Not available
Not
available
1 HGV
Dense smoke
France/Italy
1974
Mont Blanc
11,600m
Lorry
Motor
15
minutes
Not
available
Dense smoke
France/UK
1996
Channel
Tunnel
51,000m
HGV carrier
Polystyrene boxes
7 hours
Germany
2001
Dusseldorf
U Bahn
Not
available
Railway/metro train
Train roof
caught fire
Not
available
1 train
Not available
Germany
2001
Kurt Schumacher Platz

station Berlin
Not
available
Train
Arc lamp
Not
available
Not
available
Severe smoke
production
Germany
2000
Berlin
U Bahn
Not
available
Train
Not available
Not
available
1 train
Not available
Germany
2000
Saukopf
Weinheim
2,700m
Car
Not available
Not
available
1 car
Not available
Germany
1999
Candid
Munchen
252m
Car
Car engine
Not
available
1 car
Not available
Germany
1999
Leinebush
Gttingen
Not
available
High speed
cargo train
Ball bearings
overheated and
train derailed
>12
hours
1 train
Not available
Germany
1991
Bonn
U Bahn
Not
available
Train
Electrical fire broke

out on a train
in a station
Not
available
1 train
Not available
Germany
1984
Landungsbruken U Bahn
Not
available
Station fire
Arson attack
Not
available
Not
available
Severe damage
$3 million
Germany
1983
Hauptbahnhof
U Bahn Munich
Not
available
Train
Electrical fire
Not
available
Not
available
Damage
>$2 million
Germany
1981
Ramersdorf
U Bahn Bonn
Not
available
Station fire
Technical fault
Not
available
Not
available
Damage
$0.5 million
Breaks caught fire 2 hours
1 HGV carrier, Explosive spalling

10 HGVs
of concrete lining
18
Consequences
Duration
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
Germany
1980
Altora U Bahn
Hamburg
Not
available
Train
Arson attack
Not
available
Not
available
Damage
$5 million
Germany
1978
Hansaring
U Bahn
Cologne
Not
available
Train fire
Not available
Not
available
1 train
Damage
$1.2 million
Germany
1977
Berlin
U Bahn
Not
available
Station fire
Fire during
construction
Not
available
Not
available
Not available
Germany
1972
Alexanderplatz
U Bahn
East Berlin
Not
available
Train fire
Derailment
Not
available
1 train
Not available
Germany
1968
Moorfleet
Hamburg
243m
HGV carrying 14 tons

polyethylene bags
Overheated
brakes
1 hour
30 min
Not
available
Severe concrete
spalling over 34m
Hong Kong
2000
Cross
Harbour
1,800m
Car
Not available
hour
1 car
Not available
Italy
2001
Prapontin
A32 TorinoBardonecchia
4,409m
Romanian truck,
loaded with beets
Mechanical
problem
Not
available
Not
available
Closed until 6 June

in westerly direction
Italy
1999
Salerno
Railway
9,000m
Passenger train
1 train
Not available
Italy
1997
Exilles rail
Susa
2,100m
Train transporting
cars
A cars door swung

open hitting electrical
wiring causing fire
5 hours
13 freight
wagons,
156 cars
Concrete spalling
Italy
1997
Prapontin
A32 TorinoBardonecchia
4,409m
HGV
Overheated
brakes
4 hours
1 HGV
Explosive spalling
of concrete lining
Italy
1996
Isola Delle
Femmine
Palermo
148m
1 tanker with liquid

Not
gas, 1 mini small bus
available
1 tanker,
1 bus,
18 cars
Serious damage,
tunnel closed for
2 days
Italy
1993
Serra Ripoli
BologneFlorence
442m
1 car/lorry with rolls

of paper
Collision
2 hours
30 min
5 lorries,
11 cars
Little damage
Italy
1984
San
Benedetto
18,500m
Cars
Bomb attack
>2 hours
Cars
Severe damage
to structure
Italy
1983
Pecorile Galleria
Gnes Savone
662m
Lorry with fish
Not
available
10 cars
Little damage
Japan
1980
Kajiwara
740m
1 hour
30 min
One 4 ton
truck, one
10 ton truck
Serious damage
over 280m
Japan
1980
Sakai
459m
Truck
Collision
Japan
1979
Shitzuoka
Nihonzaka
2,045m
4 lorries, 2 cars
Japan
1972
Hokoriku
Fukui
Not
available
Passenger train
Restaurant car fire
Japan
1967
Suzaka
244m
Truck carrying 600

polystyrene boxes
Engine fire
>11
hours
13 trucks
Not available
Mexico
1985
Mexico City
Underground
Not
available
Metro car
Not available
Not
available
1 metro
car
Not available
Mexico
1975
Mexico City
Underground
Not
available
Train
Train collision
Not
available
Not
available
Not available
Netherlands
2001
Schiphol
Airport
Not
available
Not available
Electrical
connection box
Not
available
Not
available
Not available
19
Smoke bomb set

Not
off by football fans available
1 truck with 3600 litres Gearbox fire, collision

with side wall and
of paint in 200 cans
overturning
1 truck,
3 hours 10 vehicles
168
hours
127 lorries,
46 cars
Not
2
available carriages
Not available
Serious damage
over 1100m
Not available

Consequences
Duration
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
Netherlands
1999
Amsterdam
Underground
Not
available
Railway/metro train
Train section
Not
available
1 train
Not available
Netherlands
1978
Velsen
770m
2 HGVs and 4 cars
1 hour
30 min
4 lorries,
2 cars
Serious damage
over 30m
New
Zealand
2002
Homer
Milford
1,200m
Coach
Engine
compartment
Not
available
1 coach
Not available
Norway
2003
Floyfjell
Bergen
3,100m
Car
Car crash, burst into

flames on impact
Not
available
1 car
Tunnel lining
ignited
Norway
2000
Laerdal
24,500m
Coach
Not available
Not
available
1 coach
Not available
Norway
2000
Oslofjord
Not
available
Truck
Not available
Not
available
1 truck
Not available
Norway
2000
Rotsethhorn
1,200m
Car
Collision
Not
available
Not
available
Not available
Norway
2000
Seljestad E134
DrammenHaugesund
1,272m
The trailer truck that caused the

multiple collision had a diesel fire in
the engine room before collision
Norway
1999
Oslofjord
Not
available
Not available
Explosion during
construction
started fire
Not
available
Not
available
Not available
Norway
1996
Ekeberg
1,500m
Bus
Engine fire
2 hours
1 bus
Not available
Norway
1995
Hitra
5,600m
Mobile crane
Overheated
engine
2 hours
1 crane
Not available
Norway
1993
Hovden
Hyanger
1,290m
Motorcycle, 2 cars
hour
1 motorcycle,
2 cars
111m insulation
material destroyed
Norway
1993
Vardo
Not
available
Car
Minor fire incident
Not
available
Not
available
Not available
Norway
1990
Rldal
4,656m
V W transporter
with trailer
Engine fire
50
minutes
Not
available
Little damage
Portugal
1976
Lisbon
Underground
Not
available
Train
Electrical fire
Not
available
1 train
Damage over
$1.8 million
Russia
1991
Moscow
Underground
Not
available
Train
Electrical fire
under train
Not
available
1 train
Not available
Russia
1987
Moscow
Underground
Not
available
Train
Train
Not
available
1 train
Not available
Russia
1981
Okyabrskaya
Underground
Moscow
Not
available
Station fire
Short circuit
Not
available
Not
available
Damage
$250,000
Russia
1974
Moscow
Underground
Not
available
Station fire
Electrical fault
Not
available
Not
available
Not available
Slovenia
2004
Trojane
2,900m
Not available
Diesel powered
air compressor
Not
available
Not
available
Not available
Slovenia
2003
Golovec
Ljubljana
700m
Coach
Not available
Not
available
1 coach
Not available
Slovenia
2003
Locica
800m
HGV
Not available
Not
available
1 HGV
Not available
1 lorry,
Serious damage
45
6 cars, 1
over 1600m, tunnel
minutes motorcycle closed for day
20
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
South Africa
1994
Huguenot
3,914m
Bus with
45 passengers
Electrical fault
in gearbox
South Korea
2003
Daegu Jungangno
Underground
Station
400m
Train
Spain
2003
Guadarrama
Rail
30,000m
Train
Train accident
Spain
1975
Guadarrama
3,300m
Tanker of pine resin
Spain
1944
Torre
Not
available
Sweden
1960
Stockholm
Underground
Sweden
1955
Switzerland
Switzerland
Consequences
Duration
vehicles
installations
1 coach
Serious damage
closed for 4 days
6 carriage
trains
Severe damage
to concrete
5 hours
1 train
Not available
Tanker caught fire
2 hours
45 min
1 tanker
Severe damage to
tunnel structure and
toxic smoke
Train fire
Multi train collision
>24
hours
Multiple
trains
Not available
Not
available
Train fire
Short circuit
Not
available
1 train
carriage
Not available
Not available
Not
available
Train fire
Overheating
Not
available
1 train
carriage
Not available
2001
St. Gotthard
A2
16,918m
Lorry
Front collision of
2 lorries
>48
hours
1997
St. Gotthard
A2
16,918m
Car transporter
Overheated
engine
3 hours transporter,
1 hour
Attack using petrol 24 hours
13 lorries,
Collapse of over
4 vans, 250m of tunnel lining,
6 cars
closed for 2 months
1 car
Slight damage
8 cars
Switzerland
1997
St. Gotthard
A2
16,918m
Bus
Overheated
engine
20
minutes
1 bus
Not available
Switzerland
1994
St. Gotthard
A2
16,918m
HGV
Tyre caught fire
2 hours
1 HGV carrying
750 bicycles in
carton boxes
Severe damage over

50m to tunnel lining
Switzerland
1991
Hirschengraben
Zurich
1,300m
Train
Train car
Not
available
1 train
Not available
Switzerland
1987
Gumefens
Berne
343m
1 lorry
2 hours
2 lorries,
1 van
Slight damage
Switzerland
1984
St. Gotthard
A2
16,900m
HGV
Rolls of plastic
hour
1 HGV
Damage to
tunnel facilities
Switzerland
1976
San
Bernardino
6,600m
Bus
Bus caught fire
Not
available
1 bus
Not available
Switzerland
1972
Ltschberg
Not
available
Fire during
construction work
Not available
Not
available
Not
available
Not available
Switzerland
1941
St. Gotthard
Giorinco
Not
available
Train fire
Derailment
Not
available
1 train
Not available
Switzerland
1932
Gtschtunnel
Not
available
Train fire
Train collision
Not
available
2 trains
Not available
Switzerland
1926
Riekentunnel
Not
available
Goods train
Caught fire and

stopped in tunnel
Not
available
1 goods
train
Dense smoke
Switzerland
1969
Simplon
19,800m
Passenger train
Rear carriage
caught fire
Not
available
1 train
Not available
UK
1994
Kingsway
Liverpool
2,000m
Bus
Bus caught fire
1 hour
1 bus
Minor damage
to tunnel
UK
1987
Kings Cross
Underground
Station
Not
available
Wooden escalator
Grease and fibres

under escalator floor
6 hours
Not
available
Train station
21
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
UK
1984
Oxford Circus
Underground
Not
available
Equipment
Equipment in
maintenance tunnel
UK
1984
Summit
2,600m
Train with 13 tankers of

petroleum spirit
Derailment
Consequences
Duration
vehicles
installations
Not
available
Not
available
Equipment
destroyed
1 diesel
Severe damage
to structure
72 hours locomotive,
13 tankers
UK
1982
Picadilly Line
London
Underground
Not
available
Cable fire
Electrical cable fire
Not
available
Not
available
Not available
UK
1981
London
Underground
Not
available
Station fire
Not available
Not
available
Not
available
Severe damage
UK
1976
Finsbury Park
Underground
Not
available
Train
Cable fire
Not
available
1 train
Not available
UK
1975
Goodge Street
London
Underground
Not
available
Cross passage
Not available
Not
available
Not
available
Not available
UK
1975
Moorgate
London
Underground
Not
available
Train
Derailment, train
hitting wall
Not
available
1 train
Not available
UK
1960
Redbridge
London
Underground
Not
available
Train fire
Arcing in
receptacle box
Not
available
1 train
Dense smoke
UK
1958
Holland Park
London
Underground
Not
available
Train fire
Arcing in
receptacle box
Not
available
1 train
Dense smoke
UK
1949
Penmanshiel
Not
available
Train fire
Not available
Not
available
1 train
Not available
UK
1945
London
Underground
Not
available
Train fire
Collision
Not
available
2 trains
Not available
UK
1905
London
Underground
Not
available
Train fire
Not available
Not
available
1 train
Not available
UK
1866
Welwyn
North
Not
available
Goods train
Collision
Not
available
3 trains
Not available
USA
2007
San Francisco
McArthur
bridge
Not
available
Petrol tanker
Caught fire
Not
available
1 tanker
Bridge deck
collapsed
USA
2002
Ted Williams
2,600m
Coach
Electrical
compartment
Not
available
1 coach
Severe smoke
production
USA
2001
Howard street
Baltimore
2,253m
Cargo train
Emergency
brakes
12 hours
60 train cars, 8
carrying hydrochloric acid
Not available
USA
2000
New York City

Subway
Not
available
Train
Electrical supply
>2 hours
1 train
Not available
USA
1999
New York City

Subway
Not
available
Rubbish fire
Electrical cabling
Not
available
Not
available
Not available
USA
1996
Washington
DC Subway
Not
available
Railway/metro train
Short circuit on a
carriage led to
explosion and fire
Not
available
1 train
Not available
USA
1992
New York City

Subway
Not
available
Not available
Electrical fire
on the track
Not
available
Not
available
Dense smoke
USA
1992
New York City

Subway
Not
available
Metro car
Under car fire
Not
available
1 metro
car
Not available
USA
1990
Los Angeles
Subway
Not
available
Timber supports
during construction
Timber supports
Not
available
Not
available
45m of tunnel
collapsed
22
Consequences
Duration
vehicles
installations
Country
Year
Tunnel
Length
Vehicle where
fire occurred
Most possible
cause of fire
USA
1990
New York City

Subway
Not
available
Cable fire
Cable
Not
available
USA
1985
Grand Central
Station
New York
Not
available
Station fire
Arson attack
Not
available
Not
Severe damage to
available station, $3 million
USA
1984
New York City

Subway
Not
available
Subway car
Under car fire
Not
available
1 metro
car
Not available
USA
1984
New York City

Subway
Not
available
Subway car
Cable fire
Not
available
2 trains
Dense smoke
USA
1982
Caldecott
1,028m
2h
40min
3 lorries,
1 coach,
4 cars
Serious damage
over 580m
USA
1982
New York City

Subway
Not
available
Train
Not available
6 hours
4
carriages
Not available
USA
1982
Washington
DC Subway
Not
available
Train
Derailment
Not
available
1 train
Not available
USA
1981
New York City

Subway
Not
available
Train
Fault in current
collectors led to
explosion
hour
1 train
Not available
USA
1979
Eric Street
Subway
Philadelphia
Not
available
Train fire
Transformer fire
Not
available
1 train
Not available
USA
1979
New York City

Subway
Not
available
Station fire
Discarded cigarette
ignited oil spillage
on the track
Not
available
Not
available
Dense smoke
USA
1979
San Francisco
Subway
Not
available
Train fire
Short circuit
Not
available
1 train
Severe smoke
production
USA
1978
Baltimore
Harbour
Freeway
Not
available
Truck and fuel tanker
Collision
USA
1975
Blue
Mountain
1,300m
Truck carrying fish oil
USA
1975
Boston
Subway
Not
available
USA
1975
New York City

Subway
USA
1974
USA
1 car, 1 coach, 1 lorry with

33000 litres of petrol
1 train
Dense smoke
1 truck, 1
Not
fuel tanker,
available
1 HGV
Not available
Engine fire
Not
available
1 truck
Not available
Subway car
Broken catenary
led to a fire
Not
available
1 metro
car
Not available
Not
available
Subway rail system
Technical fault
Not
available
Not
available
Not available
Chesapeake
Bay
Not
available
HGV carrying
190 litre fuel tank
Exploding tyre
4 hours
1 HGV
Not available
1974
Congress
New York
Not
available
Goods train
Derailment
Not
available
Not
available
Not available
USA
1971
Sylmar
8,000m
Gas explosion during

construction
Not available
Not
available
Not
available
Not available
USA
1970
New York City

Subway
Not
available
Train fire
Not available
Not
available
1 train
Not available
USA
1970
Wallace
1,000m
Truck fire
Engine caught fire
USA
1949
Holland tunnel
New York
2,550m
Lorry with 11 tons

of carbondisulfide
Load falling off

lorry, explosion
4 hours
Yugoslavia
1971
Wranduk
Zenica
1,500m
Train fire
Engine fire
Not
available
23
Not
1 camper
available
truck
Not available
10 lorries, Serious damage over

13 cars 200m, dense smoke
1 train
Not available
3. TYPES OF FIRE EXPOSURE

In recent years, a great deal of research has been undertaken internationally to ascertain the types
of fire which can occur in tunnels and underground spaces. This research has taken place in both
real tunnels, and under laboratory conditions. As a consequence of the data obtained from these
tests, a series of time/temperature curves for the various exposures have been developed and are
detailed below.
Whilst research in tunnel fire phenomena continues, it should be noted that that existing data indicates
that fires within tunnels show the severity to be much higher than would be experienced under open air
conditions. By comparing heat release rate (HRR) data (understood by many to be a good measure of
fire severity) from tests carried out on different vehicle types, wooden crib fires, fuel oil tray experiments
etc, and comparing the results from tests within tunnels to those with the same tests carried out in the
open air, the conclusion has been arrived at that a tunnel can increase the HRR for a given fire load by
up to four times. Further experimentation shows that the increase will vary with the ratio of the fire width
to the tunnel width in a cubic manner.
The methods of ventilating a tunnel can also have a marked effect on the HRR of the burning items,
and thus should be factored in to any proposals when designing the type and period of fire protection
being specified.
FIGURE 5: TIME/TEMPERATURE CURVES
30 min
60 min
90 min
120 min
150 min
180 min
RWS, RijksWaterStaat (The Netherlands)
RABT-ZTV for train (Germany)
HCM, Modied Hydrocarbon (France)
RABT-ZTV for car (Germany)
HC, Hydrocarbon Eurocode 1
ISO - 834, Cellulosic, Buildings
24
3.1 CELLULOSIC CURVE

Standard fire tests to which most specimens of elements of construction are subjected are based on
the use of the Cellulosic time/temperature curve, as defined in various national standards, e.g. ISO834,
ASTM119, BS476: Part 20, DIN4102 and AS1530. Although there are other types of fire test curves,
e.g. BS7436, the curve as detailed below is the lowest used in normal practice. This curve is based on
the burning rate of the materials found in general building materials and contents. In itself, the Cellulosic
curve is based upon research dating back to the very early 20th century but it is recognised that with
the use of thermoplastic and other modern materials, the Cellulosic curve could be considered less
onerous than it should be when applied to modern building design and contents.
TABLE 3: CELLULOSIC FIRE CURVE
As will be seen from TABLE 3, and the following
tables, the time period for Cellulosic fires, with
durations of tests up to six hours, is far in excess of
those for Hydrocarbon and RWS fires. However, the
much slower rise in temperature leads to much less
damage on concrete structures.
The temperature development of the cellulosic fire
curve is described by the following equation:
T = 20 + 345* Log (8*t + 1)
3.2 HYDROCARBON CURVE
Time
(minutes)
Furnace
temperature (C)
Time
(minutes)
Furnace
temperature (C)
20
90
1006
576
120
1049
10
678
150
1082
15
738
180
1110
20
781
210
1133
30
842
240
1153
45
902
300
1186
60
945
360
1214
Although the Cellulosic curve has been in use for many years, it soon became apparent that the burning
rates for certain materials, e.g. petrol gas, chemicals etc, were well in excess of the rate at which for
instance, timber would burn. As such there was a need for an alternative exposure for the purpose of
carrying out tests on structures and materials used within the petrochemical industry. Thus the
hydrocarbon curve was developed. Initially, this time/temperature curve was developed separately by
various gas and oil companies. All had slight differences.However, today, the curve as detailed in FIGURE
5 reflects the relationship between time and temperature generally used in contemporary testing.
The hydrocarbon curve is applicable where small petroleum type fires might occur, e.g. car fuel tanks,
petrol or oil tankers, certain chemical tankers. In fact, although the hydrocarbon curve is based on a
standardised type fire, there are numerous types of fire associated with petrochemical fuels, some of
which are detailed below:
CLOUD FIRE A transient fire resulting form the ignition of a cloud of gas or vapour and not subject to
significant flame acceleration via the effects of confinement or turbulence. It can therefore only
occur after a relatively slow release of hydrocarbon and in an open, free space.
FIREBALL The rapid turbulent combustion of fuel as an expanding, usually rising ball of flame. It is
more intense than a cloud fire and can be close to an explosion.
BLEVE A Boiling Liquid Expanding Vapour Explosion which results from the sudden failure of a vessel
containing a pressurised liquid at a temperature well above its normal (atmospheric ) boiling point,
e.g. a LPG tanker.
POOL FIRE A turbulent diffusion fire burning above a horizontal pool of vapourising fuel under conditions
where the fuel vapour of gas has zero or little initial momentum. A burning pool fire is extremely difficult
to control. It may accompany a jet fire where burning liquid is spilling from the jet stream.
RUNNING FIRE A fire from a burning liquid which flows by gravity over surfaces, such as following the
slope or camber of a road tunnel.
JET/SPRAY FIRE A turbulent diffusion flame resulting from the combustion of a fuel continuously
released with some significant momentum in a particular direction.
TABLE 4: POTENTIAL FIRE DURATION
Fire types
Potential duration
Cellulosic fire
Hours
Hydrocarbon fires are different from Cellulosic in the

manner in which the temperature increase is far more
rapid and that after the initial 30 minute rise, the
temperature follows an almost straight horizontal line.
However, it should be noted that the peak
temperature of a hydrocarbon curve can be below
that of the cellulosic fire. Again, it should be reiterated
that it is the rapidity of the rise in temperature that
poses the greatest risk to a tunnel structure.
Cloud fire
Seconds
Fireball/BLEVE
Seconds
Pool fire
Hours
Running fire
Hours
Spray fire
Hours
The temperature development of the hydrocarbon

fire curve is described by the following equation:
Jet fire
Hours
T = 20 + 1080* (1 - 0.325*e-0.167t - 0.675*e-2.5t)

25
Hydrocarbon fires

Time
(minutes)
Furnace
temperature (C)
Time
(minutes)
Furnace
temperature (C)
887
60
1100
948
90
1100
10
1034
120
1100
30
1098
120+
1100
TABLE 5: HYDROCARBON FIRE CURVE

Time
(minutes)
Furnace
temperature (C)
Time
(minutes)
Furnace
temperature (C)
1047
60
1300
1120
90
1300
10
1222
120
1300
30
1297
The figures given in TABLE 5 referring to hydrocarbon fire

temperatures should not be confused with those relating to
the modified hydrocarbon curve (TABLE 6) which is now in use
in some countries. This modified hydrocarbon curve has a
temperature rise similar to that of the RABT, but with a higher
maximum temperature, reaching 1300C, only slightly under
that achieved using the RWS curve. This modified
Hydrocarbon exposure is then part way between RWS and
RABT requirements and is much more severe than exposure
to the standard hydrocarbon curve detailed within such
standards as UL1709, BS476: Part 20: Appendix D etc.
The temperature development of the modified hydrocarbon
fire curve is described by the following equation:
T = 20 + 1280* (1 - 0.325*e-0.167t - 0.675*e-2.5t)
3.3 RABT ZTV CURVE
The RABT curve was developed in Germany as a result of a

series of tunnel fire test programmes such as the Eureka
project. In the RABT curve, the temperature rise is very
TABLE 6: HCM FIRE CURVE
rapid up to 1200C within 5 minutes, faster than the
Hydrocarbon curve which rises only to 1100C after 60 minutes. The duration of the 1200C exposure
is shorter than other curves with the temperature drop off starting to occur at 30 or 60 minutes, see
FIGURE 5 on page 24.
120+
1300
The RABT test curve can be adapted to meet specific

requirements. In testing to this exposure, the heat rise is very
rapid, but is only held for a period of 30 minutes, similar to the
sort of temperature rise expected from a single truck fire, but
with a cooling down period of 110 minutes. If required, for
specific types of exposure, the heating period can be extended
to 60 minutes or more, but the 110 minute cooling period
would still be applied. The inclusion of the controlled cooling
period after the 30 and 60 minute period or heating is very
important, as the cooling process can often lead to rapid
deterioration of the concrete or any protection system.
Time
(minutes)
Furnace
temperature (C)
15
1200
30
1200
140
15
TABLE 7: RABT FIRE CURVE
3.4 RWS (RIJKSWATERSTAAT) CURVE

The RWS curve was developed by the Rijkswaterstaat in the Netherlands. This curve is based on the
assumption that in a worst case scenario, a fuel oil or petrol tanker with a fire load of 300MW lasting up
to 120 minutes could occur. The RWS curve was based on the results of testing carried out by TNO in
the Netherlands in 1979. The difference between the RWS and the Hydrocarbon curve, bearing in mind
that they both use similar fire load materials, is that the latter is based on the temperatures that would
be expected from a fire occurring within a relatively open space. Where some dissipation of the heat
occurs, however, the RWS curve is based on the level of temperature expected when a fire occurs in
an enclosed area, such as a tunnel, where there is little or no chance of heat dissipating into the
surrounding atmosphere. The RWS curve simulates the initial rapid growth of a fire using a petroleum
tanker as the source, and the gradual drop in temperatures to be expected as the fuel load is burnt off.
In the Netherlands, the RWS curve is applied for durations up to 120 minutes, at which time it is
assumed the fire load has reduced sufficiently for fire fighting personnel to be able to gain access to the
source and start in their attempts to extinguish the fire. However, in Switzerland, where tunnels through
mountains tend to be far longer in length and more remote in their location, the RWS curve is also
applied, but often extended to 180 minutes exposure. The failure criteria for specimens exposed to the
RWS time/temperature curve is that the temperature of the interface between the soffit of the concrete
TABLE 8: RWS FIRE CURVE
and the protective covering should not exceed 380C and the temperature on the reinforcement should
not exceed 250C. For high strength concrete, the interface
Time
Furnace
Time
Furnace
temperature is often reduced to a maximum of 250C.
(minutes)
temperature (C)
(minutes)
temperature (C)
In the context of a European research programme on tunnel
safety, comprehensive large scale tests were carried out in
3
890
60
1350
the abandoned Runehamar road tunnel in the western part
5
1140
90
1300
of Norway in September 2003. Semi-trailer fires, similar in
size to recent fires in the Mont-Blanc-Tunnel (France/Italy)
10
1200
120
1200
and St. Gotthard Tunnel (Switzerland), were a particular
30
1300
120+
1200
consideration. The Runehamar tests were conducted by the
Swedish National Testing and Research Institute (SP) in
collaboration with their UPTUN partners: TNO Building and Construction Research from the
Netherlands and the Norwegian Fire Research Laboratory (SINTEF/NBL).
26
TESTS & GAS TEMPERATURES AN
PROJECT
Taken from Large scale fire test in Runehamar Tunnel in Norway
In total, four tests were performed on a fire in a semi-trailer set-up. In three tests mixtures of different
cellulose and plastic materials were used to simulate the fire load, and in one test a real commodity
consisting of furniture and fixtures was used. In all tests the mass ratio was approximately 80% cellulose to
20% plastic. A polyester tarpaulin covered the cargo. The commodities are described in more detail below.
TABLE 9: COMMODITIES
USED AS FUEL IN
THE FOUR TESTS
The reason for using
furniture is that in the
past a test was carried
out (e.g. EUREKA 499
project)
with
similar
materials and a very high
ventilation rate of 6m/s at
the start of the test. This
particular test provides a
good point of comparison
between the data from
the Runehamar tests
and the EUREKA tests.
The commodities were
placed on particle board
in a storage rack system
(see FIGURES 6, 7 and 8
below) to simulate a
semi-trailer measuring
10450mm by 2900mm.
The total height was
4500mm. The height of
the platform from the floor
was 1100mm.
TEST #
Description of fire load
Target
Total
weight
Theoretical
calorific
energy
Mass
ratio of
plastic
360 wood pallets measuring 1200mm x

800mm x 150mm, 20 wood pallets
measuring 1200mm x 1000mm x 150mm
and 74 PE plastic pallets measuring
1200mm x 800mm x 150mm.
32 wood
pallets and
6 PE pallets
10,911kg
240GJ
18%
216 wood pallets and 240 PUR

mattresses measuring 1200mm x 800mm
x 150mm.
20 wood
pallets and
20 PUR
mattresses
6,853kg
129GJ
18%
Furniture and fixtures (tightly packed

plastic and wood cabinet doors,
upholstered PUR arm rest, upholstered
sofas, stuffed animals, potted (plastic)
plant, toy house of wood, plastic toys), 10
large rubber tyres (800kg).
Upholstered
sofa and
arm rest
8,500kg
152GJ
18%
(tyres not
included)
600 corrugated paper cartons with

interiors (600mm x 400mm x 500mm;
L x W x H) and 15% of total mass of
unexpanded polystyrene (PS) cups
(18,000 cups) and 40 wood pallets
(1200mm x 1000mm x 150mm).
4 wood pallets
and 40 cartons
with PS cups
(1,800 cups)
3,120kg
67GJ
19%
FIGURE 6: SET-UP FOR TEST #1 (WOOD AND PLASTIC PALLETS)
FIGURE 7: SET-UP FOR TEST #2 (WOOD PALLETS AND PUR MATTRESSES)
FIGURE 8: SET-UP FOR TEST #3 (WOOD FLAT

PACK FURNITURE AND PLASTIC TOYS)
27

TESTS & GAS TEMPERATURES AN
PROJECT
Taken from Large scale fire test in Runehamar Tunnel in Norway
FIGURE 9: SET-UP FOR TEST #4 (PLASTIC CUPS IN CARDBOARD BOXES ON WOOD PALLETS)
TEST #1
TEST #2
TEST #3
TEST #4
The test fire was located 560m 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 test fire is shown in FIGURE 10. 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 ignition the tarpaulin was replaced. At a distance
of 15m from the downstream side of the test site there was a target consisting of one stack of the same
materials combination as used in the
FIGURE 10: CROSS SECTION OF THE TUNNEL AT THE TEST LOCATION
main test. This target was used to
ascertain fire spread due to radiation
and convection.
The materials used in the tests (see
FIGURE 11) were chosen to give
different fire development and
maximum heat release rates. TEST #1
with wood pallets and plastic pallets
had the highest total energy content
and gave the highest maximum heat
release rate (see FIGURE 12A). The large
amount of combustible material
also gave a longer period of elevated
gas temperatures, with the highest
maximum temperature of 1365C.
In FIGURE 12B the gas temperature near the ceiling in
TEST #1 (at 10m from the heat source) 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. The temperature then increases even
further and more rapidly than the hydrocarbon curve.
It instead follows the RWS curve, again almost
exactly, apart from the slight time variations and for a
period around 20 minutes after ignition where the
measured temperature is higher than the peak of the
RWS curve which is 1350C.
FIGURE 11 (above):
OVERVIEW OF FIRE
DEVELOPMENT
AFTER FIVE MINUTES
FIGURE 12A: MEASURED GAS TEMPERATURES

CLOSE TO THE FIRE DURING THE FOUR TESTS
The RWS curve was developed on the
assumption that a tanker fire with petrol or fuel
oil lasting for 120 minutes would give a heat
release rate of 300MW. The heat release rate
in the tests in the Runehamar tunnel did not
reach 300MW, but the temperatures recorded
still followed the RWS curve closely.
In TEST #4 only 3120kg of cardboard boxes
and polystyrene cups were used, potentially
creating the lowest calorific energy output of all
tests. However temperatures were recorded to
be in the same magnitude of TEST #1, although
for a much shorter duration.
FIGURE 12B: GAS TEMPERATURES IN TEST #1 COMPARED

WITH FOUR DIFFERENT STANDARD FIRE CURVES
28
3.5 HEAT RELEASE RATE vs TIME-TEMPERATURE CURVE

In the fire related specifications of tunnel projects, the performance of the fire protective lining, and other
passive and active fire protection measures, are described.
Passive fire protection systems are first of all based on a design fire curve in terms of temperature
development over time as the thermal attack to the system. Also, there are the thermal failure criteria of
the structure or system that requires protection, described as maximum exposure temperatures to
certain elements of the structure. The required thermal protection can be selected using these
parameters.
It is therefore imperative to prescribe the selected
design fire curve in the fire specifications of the
tunnel project, along with the thermal failure criteria.
The thermal failure criteria can sometimes be derived
from fire testing procedures and standards.
In some cases only the Heat Release Rate (HRR),
along with the fire duration, is mentioned in the
specification, without any guidance as to
timetemperature development.
This raises the question of how to convert a HRR
figure to a design fire curve? For example, which fire
curve represents 100MW for 4 hours?
In fact, there is no physical relation between HRR
and time-temperature. Therefore the question can
not be answered, due to following reasons:
3.5.1 LOCATION OF FIRE IN THE TUNNEL
Traffic type
Fire exposure
period
Representative
nominal fire curve
Pedestrian
None
None (negligible)
Bicycle
2 minutes
None (negligible)
Hay wagon
90-120 minutes
Hydrocarbon
Car (5-10MW)
30-60 minutes
Cellulosic/Hydrocarbon
Container/shuttle
120 (+) minutes
Hydrocarbon/RABT
HGV/lorry (200MW)
120 (+) minutes
RWS
120 minutes
RWS and/or
Hydrocarbon
Tanker (300MW)
240 minutes
Bus (50MW)
90-120 minutes
Hydrocarbon
MTR/light rail (40MW)
120 minutes
RWS/Hydrocarbon
Train
120 minutes
RWS
240 minutes
Hydrocarbon
If the fire is located near the entrance or exit of

the tunnel, the heat can escape from the tunnel TABLE 10: FIRE LOAD/FIRE CURVES
and dissipate into the surrounding atmosphere.
Should the fire be located in the centre of the tunnel, the heat is trapped and will start to warm the
walls and ceiling, which in turn will radiate heat back into the tunnel.
This is also related to the length of the tunnel. In a short tunnel the heat will quickly find its way to
one of the exits, decreasing the temperature in the tunnel, conversely in longer tunnels this is not
the case.
3.5.2 VENTILATION SPEED
Ventilation systems in tunnels are an important part of a holistic fire safety concept. Full scale fire
tests have shown that an increased ventilation speed in the tunnel will most likely increase the fire
size and can potentially induce fire spread from one vehicle to the other. By increasing the ventilation
speed additional oxygen is fed to the fire source, again increasing the fire size and accelerating the
consumption of the fuel, thus decreasing the duration of the fire. A slower ventilation speed reduces
the fire size but the duration of the fire will be prolonged.
Depending on the ventilation approach, the fire size and fire duration will be influenced.
Ventilation speed also influences the gas temperature in the tunnel. For a given fire size and fire
duration, increasing the ventilation speed will decrease the gas temperature away from the seat of
the fire. In this case the overall effect of increasing the ventilation speed may be a lower thermal
attack to the structure.
However, increasing the ventilation speed can result in the spread of fire from one vehicle to another.
For example, see FIGURE 53 on page 54.
It is therefore very difficult to ascertain whether or not high or low ventilation speeds will increase or
decrease temperatures and adversely effect fire spread.
29
3.5.3 GRADIENT
A gradient or slope of the tunnel influences the so-called Chimney Effect. If the tunnel has no
gradient the heat and smoke will spread through the tunnel in the same direction as the ventilation
direction. Where a low ventilation speed exists, the heat will build up at the location of the fire,
leading to an increase in temperature.
With a gradient of 5%, for example, the heat and smoke will climb upwards. If the ventilation goes
in the same direction, the heat will be taken from the fire location more rapidly, reducing temperature
development at the fire source.
3.5.4 CROSS SECTIONAL AREA
A fire in a large tunnel will build up less heat as opposed to a fire in a tunnel with a smaller cross
sectional area. In a large tunnel, the volume of air that needs to be heated is greater and also the
surface area of the walls and ceilings is larger, and are therefore able to absorb more heat.
3.5.5 TIME/DURATION OF THE FIRE
Depending on the volume and type of combustible materials involved in the fire, duration will be
influenced, along with the temperature rise in the first minutes of ignition.
The illustrations at left depict the major difference between fires where heat can dissipate, as in
most buildings, and those in tunnels where the heat is trapped and creates a chimney effect.
If all the above parameters are known, a Computational Fluid Dynamics (CFD) calculation can be
made, demonstrating an understanding of temperature development in certain fire scenarios. This can
be combined with a finite element analysis (FEA) in order to ascertain temperatures on elements of
the structure.
However, tunnels are designed to cope with several fire scenarios which makes it unfeasible to derive
a time-temperature curve from each fire scenario.
The above implies that the Heat Release Rate is an important parameter for fire safety design in tunnels,
in fact, amongst other items, ventilation systems are designed using Heat Release Rate data.
For the design of passive fire protection systems such as the protection of the structural lining, however,
the Heat Release Rate data cannot and should not be utilised as the sole determining factor.
Top:
HEAT IN A BUILDING
DISSIPATES
Bottom:
HEAT TRAPPED
IN A TUNNEL CREATES
A CHIMNEY EFFECT
3.6 FIRE PROTECTION REQUIREMENTS OVERVIEW

Throughout the world, fire development in a tunnel will be the same under the exact same
circumstances. In other words, how would the fire know in which country the tunnel is located?
The same goes for the structural lining and other materials that should survive a tunnel fire, e.g. cables,
fire doors. The same material will behave in the same manner under the exact same fire circumstances,
regardless of the country it is in.
In light of this commonality, a degree of harmonisation of fire protection requirements for tunnels might
be a reasonable expectation. However, TABLE 11 (on opposite page) indicates that there are still
substantial differences in requirements, in terms of design fire curves and thermal failure criteria for
concrete protection.
Although the EU research programmes (UPTUN, DARTS, FIT etc.) have contributed a lot to
harmonisation, there is still much work to conduct in this respect.
30
TABLE 11: OVERVIEW OF FIRE PROTECTION REQUIREMENTS

Country
Netherlands
USA
Code/Standard Traffic type
RWS
1998-CVBR1161 (rev. 1)
NFPA 502
Road
Fire curve
RWS
Construction
method
Immersed,
cut and
cover (C&C)
Bored or
drilled
Road
RWS
Immersed,
cut and
cover (C&C)
Bored or
drilled
Concrete type
Cast in
place
Prefabricated
Cast in
place
Prefabricated
N0: none
CETU
Road
N2: HCM 2 hours
T interface <380C
T rebar <250C at
25mm concrete cover
No spalling
T interface <380C
T rebar <250C at
25mm concrete cover
No spalling
T interface <380C
N1: ISO 2hours

France
Temperature criteria
Any
Any
N3: ISO 4 hours

and HCM 2 hours
T rebar <250C
T unexposed face
<60C
RABT
Road
RABT-30+110 cooling down
Any
Any
T rebar <300C
ZTV/EBA
Rail
RABT-60/90+110
Any
Any
T rebar <300C
Germany
T1: T rebar average
<200C, max. <250C
T1: T interface average
<330C, max. <380C
Italy
UNI 11076
All
RWS
Any
Any
T2: T rebar average

<250C, max. <290C
<380C, max. <420C
T3: T rebar average
<300C, max. <350C
<430C, max. <460C
T interface <350C
Austria
OVBB
Any
RWS, 3 hours
Any
Any
Singapore
LTAS
Road,
KPE Tunnel
RWS
Immersed
Cast in place
L > 1500m;
RABT 120+110
GB 50016
Hazardous
goods
China
GB 50016
Nonhazardous
goods
L > 500 but

1500m;
RABT 90+110
31
Road,
Palm Jumeirah
Tunnel
T interface <380C
T rebar <250C
RABT: T interface <380C
Any
Any
RABT: T rebar <300C at

25mm concrete cover
HC: T interface <380C
L 500m;
HC, 2 hours
HC: T rebar <250C at

25mm concrete cover
L > 3000m;
RABT 120+110
RABT: T interface <380C
L > 1500 but

3000m;
RABT 90+110
Any
Any
RWS
RABT: T rebar <300C at

25mm concrete cover
HC: T interface <380C
HC: T rebar <250C at
25mm concrete cover
L > 500 but

1500m; HC, 2 hours
UAE, Dubai
T rebar <250C at
40mm concrete cover
Cut and cover

(C&C)
Cast in place
T interface <380C
T rebar <250C
Source: Efectis BV The Netherlands
Spalling is an umbrella term, covering different damage phenomena that may occur to a concrete
structure during fire. These phenomena are caused by different mechanisms: pore pressure, thermal
gradient, internal thermal micro-cracking, cracking around reinforcement bars and strength loss
due to chemical transitions. In different combinations of these mechanisms, possible spalling
phenomena include violent spalling, progressive gradual spalling, explosive spalling, corner spalling and
post cooling spalling.
Spalling of concrete during fire causes serious damage to concrete structures, with significant economic
costs and risk to human life. New developments in concrete technology such as improved grain size
distribution and the application of extra fine particles have resulted in concrete types with improved
durability, strength and workability. However, these high performance concrete types have been shown
to be more susceptible to spalling during fire than ordinary concrete types. The problem of spalling in
buildings has been known for decades and further highlighted by recent intense tunnel fires in Europe.
As a consequence of severe damage due to spalling and the non-operational time of tunnels after a fire,
the fire resistance of newly developed concrete types has been called into question.
4.1 HEATING RATE AND INTERNAL STRESSES

During a tunnel fire, air temperatures can rise to over 1300C within just a few minutes. Compared to
building fires, this is a much more severe situation, giving a large thermal shock to the structure. For the
design of buildings, there is worldwide agreement on the use of the ISO-834 standard fire, which
prescribes a slower temperature development, as shown in FIGURE 5 on page 24. For tunnels many
design curves are available. A few of these fire curves are also shown in FIGURE 5. Although usually a
less expensive solution is obtained by using a lower fire curve, this may well lead to unsafe situations.
Recent full scale tunnel fire tests carried out by the UPTUN consortium, for example, have shown that
fire temperatures may quickly reach 1300C to 1400C. This is a critical issue because many insulation
materials cannot withstand temperatures above 1200C or may be unable to withstand the thermal
shock of such a rapidly developing fire, and may therefore be unsuitable for protection of a tunnel lining.
During heating, stresses develop inside the concrete cross-section. Thermal gradients and moisture
pressure lead to mechanical stresses that may cause internal and external cracking as well as spalling
of concrete.
4.2 SPALLING OF CONCRETE

Spalling of concrete is one cause of damage to the structure. Other causes of damage that develop
during fire exposure are internal cracking, irreversible plastic and creep strains and chemical transitions.
These forms of damage might eventually lead to collapse due to a failure mechanism like bending,
shear, anchorage or buckling.
Often when concrete is damaged in a real fire the damage is called spalling. In many cases this is not
correct; other failure mechanisms such as shear failure can also lead to severely damaged concrete.
Real spalling can occur in different forms, each of which is caused by a specific combination of the
following mechanisms:
Pore pressure rises due to evaporating water as the temperature rises;
Compression of the heated surface due to a thermal gradient in the cross section;
Internal cracking due to differences in thermal expansion between aggregate and cement mix;
Cracking due to differences in thermal expansion/deformation between concrete and
reinforcement bars;
Strength loss due to chemical transitions during heating.
The mechanisms act on different scales:
MACRO-LEVEL Concrete considered as a grey homogeneous material with uniformly distributed
material properties. On this level, the thermal stresses that result from the thermal gradients over
the cross section must be considered, taking into account the actual geometry, support conditions
and loading configuration.
MESO-LEVEL Concrete considered as a mix of aggregate and cement mix, each with its own material
properties. On this level, the cracking due to differential thermal expansion between aggregate,
mortar and reinforcement must be considered.
MICRO-LEVEL Cement mix, aggregate particles or interface layers considered as a mix of chemical
constituents. On this level, the pore pressures and the degradation of mechanical properties due to
chemical transitions and dehydration must be considered.
32
During fire tests the observations of spalling of concrete

cover a wide range. These are, in random order: observation
of spalling with slow (1C/minute) or fast (250C/minute)
heating, from gradual to explosive spalling, cracking along
or through aggregate grains, spalling in the beginning of the
fire or after some time, stopping after some time or
progressing, stopping at the reinforcement level or
continuing far beyond it, and so on. The different observed
spalling phenomena are described below, including their
relationship to the previously mentioned mechanisms (see
also Breunese & Fellinger, 2003).
A summary of these relationships is given in TABLE 12 at right.
Pose pressure
due to
evaporation of
moisture
Violent spalling
Internal cracking
Cracking
due to different due to different
Compression
thermal
due to thermal thermal expansion
of aggregatedeformation of
gradient
cement paste
concrete-steel
Explosive
spalling
Corner spalling
Post-cooling
spalling
4.2.1 VIOLENT SPALLING

Violent spalling is the separation of small or larger pieces of concrete from the cross section, during
which energy is released in the form of pieces and small slices of concrete popping off with a certain
speed, and also a popping or cracking sound. This type of spalling is caused by pore pressure and
thermal gradients. Internal cracking on the meso-level also influences this spalling process. The
surface compression during heating can increase due to lateral restraint, reinforcement,
prestressing, large concrete thickness and a high heating rate. Pore pressures are dependent on
heating rate, moisture content, permeability, porosity and the presence of polypropylene fibres
(artificial permeability). Furthermore, an increased ductility of concrete by the addition of steel fibres
has sometimes been reported to reduce the risk of this type of spalling. (Fellinger & Both, 1997)
4.2.2 PROGRESSIVE GRADUAL SPALLING (SLOUGHING OFF)
Sloughing off is the form of spalling that is caused by strength loss due to internal cracking (mesolevel) and chemical deterioration of the cement mix (micro-level). This type of spalling is related to
the attained temperature of the concrete (instead of heating rate). If the concrete is heated to a very
high temperature the strength will be too low to carry its own weight, causing small pieces of
concrete to fall down without much sound. This type of spalling is likely to occur on a slab heated
from below, since gravity will force the cracked pieces of concrete from the cross section.
4.2.3 CORNER SPALLING
Corner spalling is the type of spalling that occurs when a corner of concrete breaks off at the
location of a reinforcement bar. Inhomogeneous heating of concrete leads to a deformation
(ovalisation) of the concrete around the uniformly heated reinforcement bar. This difference in
deformation causes splitting stresses in the concrete, leading to splitting cracks that can cause the
corner of a column or slab to break off.
4.2.4 EXPLOSIVE SPALLING
Explosive spalling is the result of a combination of rising pore pressures and thermal gradients in
the cross-section. At the front of heat penetration, a moisture clog (an area with high pore
pressure) develops inside the concrete. Part of the moisture is pushed further into the colder part
of the concrete due to the pressure gradient at the back of the clog. If the heated surface is under
compression due to a thermal gradient, the complete heated surface may explode away with a loud
bang. This type of spalling is especially likely to occur on structural members heated from more than
one side, such as columns and beams. When moisture clogs are advancing into the concrete from
all heated sides, at some point in time the moisture clogs will meet in the centre of the crosssection, creating a sudden rise in pore pressure which may cause large parts of the cross-section
to explode. This type of spalling can also occur after a considerable duration of the fire even if the
concrete surface has been protected with an insulating layer. (Both, 1999)
4.2.5 POST-COOLING SPALLING
Post cooling spalling occurs after the fire is over, after cooling down or maybe even during
extinguishing (Khoury, 2003). This type of spalling was observed with concrete types containing
calcareous aggregate. An explanation is the rehydration of CaO to Ca(OH)2 after cooling, with an
expansion of over 40% occurs after cooling down, when moisture is again present on the concrete
surface. The expansion due to rehydration causes severe internal cracking on the meso-level and
thus complete strength loss of the concrete. Pieces of concrete keep falling down as long as there
is water to rehydrate the CaO in the dehydrated zone.
33
Sloughing off
Strength loss
due to chemical
transitions
TABLE 12:
IMPORTANT RELATIONS
BETWEEN MECHANISMS AND
SPALLING PHENOMENA
Source: Efectis BV The Netherlands
4.3 TESTING OF SPALLING BEHAVIOUR

For a spalling test, it is of great importance to simulate the practical situation as
closely as possible in the test setup. Only in this way is it possible to draw conclusions
from the test; extrapolation of test results is difficult at best. Due to the variety in
spalling test results, a test should always be performed twice in an identical lay-out.
See FIGURE 13 at left.
4.3.1 GEOMETRY, PRESTRESSING, CONCRETE MIX AND MOISTURE LEVEL
For the concrete it is important to use the concrete mix and geometry as will be used
on the project. The case of pre-cast circular segments, preferably segments made in
the factory should be used. For spalling, the prestressing level is important, and
should resemble the actual situation. The moisture level of the concrete should be at
least as high as it will be in the actual situation. In general, a specimen with higher
moisture content is more likely to spall and therefore give a safer test result.
FIGURE 13:
TEST SET UP FOR A FULL
SIZE TUNNEL SEGMENT
4.3.2 AGE OF THE SPECIMEN

The specimen must be old enough to have a moisture content close to the actual situation as if it
had been used for many years. This is necessary because spalling is strongly influenced by free
water content, porosity and permeability. After 28 days much of the final strength of concrete has
been reached, but permeability is still decreasing. For practical reasons it is of course impossible to
test segments of many years age. At TNO Netherlands, for example, the usual age of specimens
at the time of testing is at least 90 days.
4.3.3 FURNACE TEMPERATURE
The fire test must be carried out according to a suitable fire curve. It is important to achieve the
steep increase in the first 5 to 10 minutes of the test because this gives a high thermal shock to the
concrete. It is also important to achieve a sufficiently high maximum temperature because many
insulation materials may melt around 1200C.
4.3.4 INSULATION MATERIAL
If a protective layer, such as board material or cementitious spray, is used, it is important to pay
attention to the method of fixing the material to the concrete surface. The details are also extremely
important. These details include covering of hollow spaces in the concrete surface, and sufficiently
protecting objects that are fixed to the concrete.For example, a road sign fixed to the tunnel ceiling
with steel bolts in fact forms a penetration of the protection layer and may locally introduce heat into
the concrete, leading to possible spalling. Once spalling starts in such a small region, pieces of
spalling concrete may rapidly push away the remaining protection material and leave the whole
surface unprotected. For the material of the protective layer, a low moisture content during the test
is recommended. This reduces the insulation capacity of the material and thus gives a safe test. The
layer thickness should be identical in both tests. Interpolation of layer thicknesses is impossible for
spalling tests!
The latest investigations into the fire performance of concrete show that even the addition of
polypropylene fibres into the concrete mix will not always suffice to reduce water vapour pressure,
and thus can have little effect on reducing the incidence of spalling. It should also be noted that the
majority of testing to date on the performance of concrete with the addition of polypropylene fibres
has been to the standard cellulosic curve, and not to the greater requirements of tunnel fire curves.
Even for these relatively low temperature rise fires, the proportion of PP fibres to concrete mixture
required is such that the concrete is often very stiff and difficult to work. It should further be noted
that use of PP fibres will result in no provision of insulation to the concrete against rapid temperature
rise, which could result in extensive internal and external cracking of the concrete, even where
spalling is alleviated. Care should therefore be taken to ensure claims for the performance of PP
fibres are substantiated by adequate evidence.
4.4 FIRE RESISTANCE OF CONCRETE

Research has shown that concrete structures suffer surface spalling as a result of high compression
stresses in the heated outermost layers and by the generation of water vapour at high pressure behind
those layers. The probability of spalling increases with compression stress and the moisture content of
the concrete. With a moisture content of over 3% of the mass, the probability of spalling is virtually
100%. Explosive spalling presents immediate risks to emergency response personnel in fire situations
and the exposure of underlying steel can result in rapid deterioration of strength and load capacity.
It should be noted that concrete can be heated slowly and spalling will not occur, or will be minimised.
However when heated rapidly, precisely the type of fire seen in tunnels where the onset of fire growth
is extremely rapid rising to very high temperatures, the permeability of the concrete and the ability of the
moisture to find its way to the surface determines the onset and severity of spalling.
Rapid rates of heating, large compressive and tensile stresses or high moisture contents (over 5% by
volume or 2% to 3% by mass of dense concrete) can lead to excessive spalling of concrete cover at
elevated temperatures, particularly for thicknesses exceeding 40-50mm. This water is not only
physically present (moisture), but also chemically bound within the concrete (hydrated water).
34
Such spalling may impair performance by exposing the reinforcement or tendons to the fire or by
reducing the cross-sectional area of concrete. Concrete types made from limestone aggregates are less
susceptible to spalling than concrete made from aggregates containing a higher proportion of silica, e.g.
flint, quartz and granites, due to their permeability. Concrete made from manufactured lightweight
aggregates suffer a lesser degree of spalling. The use of high strength concrete has been introduced as
it can reduce the necessary thickness required to obtain a certain structural performance. However,
high strength concrete is particularly prone to very severe spalling when exposed to fire. As the
thickness of the concrete has already been reduced due to its higher strength, the effects of spalling
are even more severe than usual.
The latest investigations into alternative methods of protecting concrete against spalling show that the
incorporation of fine denier engineered fibres of polypropylene or steel into concrete will when added
in specific volumes and distributed uniformly reduce the risk of tensile forces causing explosive failure
to the parent concrete when exposed to the most rigorous fire. This is achieved by a series of pressure
relief pores occuring upon exposure to heat can reduce the effects of spalling.
The addition of polypropylene or steel fibres to the concrete require an increased amount of plasticiser,
the addition of air entraining agents in order to stabilise the concrete and retardants to prolong the
concretes opening time during its application. Results from fire research tests showed that for both
macro synthetic and steel fibre reinforced concrete, the addition of polypropylene or steel fibres, with a
recipe adjusted for this addition, spalling caused by fire can be minimised. It was also found that the
addition of polypropylene or steel fibres alone is not enough to minimise spalling as is often presumed.
It should be noted very clearly that the precise concrete recipe and the amount of polypropylene or steel
fibres is of great importance.
Based on the opinions of a number of researchers, the use of polypropylene or steel fibres for any
specific project should be carefully considered, and the specific concrete mix being used in the project
subjected to fire testing to ensure the proposed type, dimension and quantity of fibres will provide the
requisite fire performance.
It should also be noted that the majority of testing up to the present time on the performance of
concrete with the addition of polypropylene fibres has been to the standard ISO cellulosic curve, with a
small number of tests performed to the standard Hydrocarbon curve (1100C) and the RWS curve
(1350C). In addition consideration must be given to the fact that not all polypropylene or steel fibres
can be considered equal. The best fibres are those manufactured to a dimension and tolerance and
from new materials specifically for the task at hand. Testing has shown that the use of recycled plastic
as polypropylene fibres has less effect than purpose-made materials. Care should therefore be taken to
ensure claims for the performance of any particular polypropylene or steel fibres are substantiated by
adequate evidence of their performance under rapid growth, high temperature fire.
TABLE 13: FACTORS TO CONSIDER WHEN COMPARING PROMATECT-H OR PROMATECT-T BOARDS
TO POLYPROPYLENE FIBRES (PPF) FOR FIRE PROTECTION OF CONCRETE STRUCTURES
Item
PROMATECT-H or PROMATECT-T boards,

cementitious sprays
Insulation of the rebars
Fully insulates pre-set and pre-designed limits of

maximum requirements.
No insulation at all.
Bond between steel and concrete
Maximum temperatures not exceeded at reinforcement, maintains bond.
At 300C the bond between rebars and concrete will be significantly reduced.
Replacement of concrete after fire
Only the boards or sprays must be replaced after

severe fires.
All concrete, which was exposed to temperatures exceeding 300C must be

replaced. Even after small fires (T>160C) repairs are required because the
fibres have melted, and therefore can no longer fulfil their intended function.
Long term durability, chlorides
Boards and sprays have no adverse effect on

the durability of concrete structures.
PP fibres create small channels in the concrete, due to the hydrophilic

properties, enabling chlorides and sulphates to penetrate the concrete and
attack the rebars.
Relatively small damaged area.
Damaged area (T>300C) bigger than that directly affected by fire.
Avoid spalling
Boards and sprays are designed to reduce

temperature increase on and in the concrete
structure and therefore prevent spalling occurring.
Until now all tests with PP fibres have shown spalling of the concrete
specimen. PPF do not stop structural damage occurring due to high
temperatures (micro cracks can occur at 150C).
Influence on the concrete properties

the properties of concrete structures.
PPF reduces compressive strength leading to brittle failure. PPF will also cause
reduced pull out strength of anchors under fire conditions due to melted fibres.
Influence on workability of the concrete

the workability of concrete structures.
Workability decreases with increasing concentration of fibres. 3kg/m3 of fibres

dehydrates the concrete mixture severely, making it difficult to pump or pour
the concrete.
Being able to withstand all types of fires
Boards and sprays can withstand all types of

fires, up to the most severe RWS fire.
A smouldering fire will cause dehydration of the outer layer of concrete,

causing even more aggressive spalling when temperatures increase after 2030 minutes.
Influence on the clearance of

the tunnel cross section
Board and spray systems are relatively thin,

<40mm depending on the fire requirements.
PPF tunnels require larger cross sections, bigger TBMs, more tubing
segments, bigger volumes of excavated soil etc. Sacrificial linings, containing
PPF can be over 250mm thick.
Quality control of
the fire protective system
Boards and sprays are produced to ISO9001

quality standards.
How controllable is a homogeneous mix made up on site? PP fibres must be

evenly distributed throughout a mixture. Impossible to control or to check.
Therefore performance under fire is unpredictable.
Damaged area after fire
35
PP fibres
4.5 CRITERIA FOR THE FIRE RESISTANCE OF CONCRETE
Unprotected concrete with a moisture content of over 3% of the mass will suffer surface spalling in a
fire, probably after 5-30 minutes. It is also possible for aggregates in the concrete (e.g. quartz) to have
undesirable effects on its behaviour in a fire.
VALUE
The criteria for fire resistance has been drawn up by a number of official bodies. The Dutch RWS
standard suggests that, as a rule of thumb in the case of loadbearing members, account should be
taken only of cores whose temperature is less than 500C. At the tensions at which reinforcing steel is
commonly used today, steel starts to flow at 500C. In statistically determinate structures this leads to
failure. In statistically indeterminate
structures re-distribution of the
FIGURE 14: INFLUENCE OF TEMPERATURE ON CONCRETE
moments is often possible, so that
a higher temperature of the reInfluence of elevated temperatures on concrete and steel
inforcing steel need not necessarily
lead to failure.
1.2
Yield stress steel
Value at elevated temperature

Value at room temperature
1.0
New steel
Compressive stress concrete
0.8
Fresh concrete
0.6
0.4
0.2
0C
30C
55C
200C
300C
400C
500C
600C
700C
800C
900C 1000C 1100C 1200C
TEMPERATURE
Based on the requirements for

exposure to an RWS type fire:
Temperature on the concrete
interface should not exceed
380C (for bored tunnels this
limit is 200-250C).
Temperature
on
the
reinforcement should not
exceed 250C with a
minimum of 25mm concrete
cover. (Note: For exposure to
RABT, the reinforcement
temperature should not
exceed 300C.)
There is a high risk of failure due to the temperature of the steel in the concrete in columns with a high
reinforcement level under high loads. For this reason, the (non-normative) tables give a critical steel
temperature of 500C for ordinary concrete and steel and 400C for tension steel. In the Netherlands,
Rijkswaterstaat specifies for tunnels a maximum permissible concrete surface temperature of 380C.
This maximum was set not because of any perception that concrete fails at that temperature but
because it is assumed that in practice this is a temperature at which there is only a very small probability
of damage to concrete. This requirement also implies that the temperature of the underlying
reinforcement remains low, so that its strength is unimpaired.
The design of the tunnel section has an effect on fire induced collapse.
Rectangular tunnels were typically constructed using a grade C30/35 concrete. Nowadays
C40/45 is commonly used.
Failure of rectangular structures is usually due to the premature development of sagging plastic
moment caused by elevated temperatures of the concrete and the reinforcement.
FIGURE 15:
SPALLING OF CONCRETE
AFTER ACTUAL FIRE
Rectangular structures suffer from less spalling than circular tunnels and have limited
compression loads.
Circular tunnels were constructed from segmented reinforced concrete sections typically use a
C50 grade concrete or higher.
After completion, reinforcement in circular tunnels is more or less obsolete, only required to assist
handling during installation.
The reinforcement in circular tunnels is not required to take tension forces in sagging moment
because the concrete is typically in compression.
The higher strength concrete (C50) suffers a higher percentage and depth of spalling due to fine
fillers such as lime stone and fly ash, the reinforcement will however help retard the effect of
explosive spalling.
The depth of spalling under fire conditions is an average 100% deeper on these types of
circular tunnels.
FIGURE 16:
REINFORCEMENT
TEMPERATURES
36
5. CHOOSING FIRE PROTECTION MATERIALS

5.1 CONSIDERATIONS WHEN APPLYING PROTECTIVE MATERIALS
In the design of a system to protect concrete, the following questions need to be answered to determine
the correct material types to be used.
What type of fire needs to be resisted (e.g. time/temperature curve)?
How long must the protected structure survive (e.g. duration of time/temperature curve)?
Type of concrete (e.g. cast in place [immersed or cut and cover tunnels], prefabricated,
circular tunnels)?
The moisture content of the concrete?
The density of the concrete?
The aggregates used in the concrete mix itself (e.g. silicious or calcegenous)?
The period of time the structure has to be able to survive without failing and the type of fire to be
withstood, together determine the thickness of the protection that is required. The requisite protection
material thicknesses will be found in the fire test reports provided by the manufacturer of the protective
lining materials. These same reports also give guidelines for the points of attachment and the type of
fixing to be used.
5.2 BOARD MATERIALS

Board materials can be easily checked for thickness and thus the application can be guaranteed to
meet with the specifications as per the tested constructions. In addition, being mechanically fixed,
suction and wind loading from passing traffic has no adverse effect on boards, with correctly installed
products remaining in place without any deformation occurring.
Boards are completely unaffected by combustion by-products of traffic passing through tunnels, and
are also unaffected by the ingress of water. In fact, in very wet tunnels, boards can act as a conduit for
water, ensuring the excess runs off into the tunnel drainage systems rather than onto the road surface.
Board protection systems will also act as a form of filter during exposure to fire, ensuring that chlorine
and other gases given off by burning rubber and plastic used in the construction of modern vehicles,
and which are extremely corrosive in nature, do not have direct access to attack the concrete and
reinforcement of the tunnel linings.
Using a board product such as PROMATECT ensures that condensation as a result of wet tunnels
does not form on the exposed surface of the boards, but rather this small amount of moisture is
absorbed by the PROMATECT and then evaporated into the surrounding air. The absorption of water
into PROMATECT has no adverse effect on the performance of the board.
Board systems in general require little maintenance. Where access is required to periodically inspect the
concrete substrate, boards can quickly and easily be removed and reinstated, thus maintaining the fire
protection layer at all times.
Following criteria for thermal failure can be specified in order to correctly and adequately design the
required material type and thickness:
Maximum allowable interface temperature;
Maximum allowable temperature of the reinforcement, along with the cover on the reinforcement;
Maximum allowable interface heating rate (C per minute);
Maximum allowable temperature of the unexposed side of the concrete slab (in case of escape
route protection).
Depending on the specific project related requirements a combination of the above thermal design
criteria can be made.
37
6. TYPES OF TUNNEL
6.1 BORED TUNNELS USING CONCRETE SEGMENTS
A bored tunnel refers to a construction method for tunnels which involves digging a tube-like passage
through the earth. It usually refers to tunnelling through rock, as blast tunnelling is not widely used these
days. Bored tunnels are created using a full face boring machine which has cutting teeth at its front. It
creates the tunnel opening while passing waste material through to the rear. Many types of tunnel boring
cut small sections which are progressively enlarged. A full face tunnel boring machine (TBM) cuts the
complete cross section of the tunnel in one pass.
The TBM consists of a long machine with a circular cutting head that rotates against the face of the
tunnel. Attached to the cutting head is a series of steel alloy disk cutters that gouge out the rock on the
face as the machine rotates. The cutting head is pushed forward by hydraulic power. TBMs provide
several advantages over drilling and blasting. The tunnel can be bored to the exact size desired, with
smooth walls, thus eliminating the condition called overbreak, which results when explosives tear away
too much rock.
The use of TBMs also eliminates blasting accidents, noise, and earth shocks. Workers need not be
concerned with fumes or noxious gases and can clear away broken rock without stopping for blasting
intervals. A TBM can advance about 76 metres (about 250 feet) a day, depending on the diameter of
the tunnel and the type of rock being bored. Despite these advantages, TBMs have some drawbacks.
They are very costly and the cutting head must be the same diameter as that required for the tunnel.
Often the TBM is part of a long train of machines. At the rear are stored circular concrete sections,
which are installed as the TBM moves along the route, thus the tunnel is simultaneously lined as it
is drilled.
6.2 IMMERSED TUNNEL

The immersed tube is a construction method using pre-fabricated tunnel sections. While the ends are
sealed, it is lowered into position under the water and attached to other sections. It is sometimes called
a sunken tube.
Another method of underwater tunnel construction uses a caisson, or watertight chamber, made of
wood, concrete or steel. The caisson acts as a shell for the building of a foundation. The choice of one
of three types of caissons the box caisson, the open caisson or the pneumatic caisson depends
on the consistency of the earth and the circumstances of construction. Difficult conditions generally
require the use of the pneumatic caisson, in which compressed air is used to force water out of the
working chamber.
Another method of constructing underwater tunnels, such as those like the Noord tunnel in the
Netherlands, have been built by fabricating short tunnel sections in a trench in or near the riverbed or
seafloor. Each section, after completion is then sealed at the ends, floated out and located in position,
where it is then sunk onto the river or sea bed. After sinking, the sections are then attached in line by
oversized bolts to the previously sunk section. Heavy, thick concrete walls prevent the tunnel from
floating once the water is pumped from the completed sections.
6.3 CUT AND COVER (C&C) TUNNELS

A construction method which involves excavating a large trench, building a roof structure, then covering
it with earth. Commonly used for subways and in relatively flat locations.
The cut-and-cover method can also involve digging a trench, building the concrete floor, walls, and
ceiling, or installing pre-cast tunnel sections, and then refilling the trench over the tunnel. In built-up
areas in cities, use of this method is often not possible. In soft earth or mud, a large diameter pipe-like
device can be driven through the ground by jacks or compressed air. Workers remove the earth as the
pipe moves forward, its edge cutting into the earth.
38
7. FIXINGS
Source: Fischer fixing systems
Where daily use of a road or rail tunnel involves high traffic loads, the demands placed on means of
attachment will be more onerous. Traffic passing through a tunnel causes high suction loads on ceilings
due to the displacement of air by vehicles. This suction load depends on the type of vehicle (e.g. car,
train or tram) and the headroom. The value often taken is 100kg/m2. The weight of the cladding and the
number of attachment points can be determined by means of a load simulation for traffic passing
through a tunnel. The manufacturer of the pertinent cladding material should provide reports which
show the exact types of fixing methods employed, and the loads from suction etc. that the systems are
designed and tested to take. Attention needs also be paid to the material from which the fixings are
manufactured (zinc galvanised steel, stainless steel) and the condition of the concrete itself.
REFERENCES:
Therefore, before any choice of fixing can be made, consideration has to be given to three important
elements. The likely corrosion to which the anchors may be subject, the crack width if any, and the
compressive strength of the concrete.
(1)
N.N. (1989). Nichtrostende

Sthle, Eigenschaften,
Verarbeitung, Anwendung,
Normen. 2. Auflage, Verlag
Stahleisen, Dsseldorf,
S. 16/74.
(2)
Gmpel, P. (1996). Rostfreie

S t h l e , e x p e r t - v e r- l a g .
Kontakt & Stadium, Bd 493.
(3)
Hasemaier, H., beleis, A. &

Bhni, H. (1993). Korrosionsbestndige Befestigungen in
Straentunnels Feldversuche. Scheweizer Ingenieur
und Architekt Nr. 16-17,
S. 289-295.
(4)
beleis, A. & Felder, G.

(1993). Schadstoffe in Straentunnels Auswirkungen
auf die Bestndigkeit met.
Werkstoffe. VDI-Berichte Nr.
1060, S. 69-84, VDI-Verlag,
Dsseldorf.
(5)
Arnold, N., Arlt, N., Gmpel,

P. & Heimann, W. (1996).
Stainless Steels for Fasteners
for Architectural Application
in Highly Corrosive Environments Proceedings. 2nd
European Congress Stainless
Steels, Dsseldorf, 3.5.6.96,
S. 138-142.
(6)
Htterer, H., Gmpel, P. &

Ruoss, H. (1996). Schadgaskorrosion von Werkstoffen
in der Befestigungstechnik.
Materials and Corrosion, S.
355-364.
(7)
Arnold, N., Gmpel, P., Heitz,

T. & Pscheidl, P. (1997).
Chloridinduzierte Korrosion
von nicht rostenden Sthlen
in SchwimmbadhallenAtmosphren. Teil 1: Elektrolyt Magnesium-Chlorid
Materials and Corrosion 48,
S. 679/86.
7.1 LOCALISED CORROSION AND PITTING IN TUNNEL CONSTRUCTION

7.1.1 LONG SERVICE LIFE OF STAINLESS STEEL GRADES USED FOR ANCHORS
Due to their excellent long-term corrosion resistance in naturally occurring ambient conditions,
stainless steel grades are used more and more in fixing engineering.(1) and (2) Although they tend to be
more expensive, the extra costs are compensated relatively quickly by their longer service life and
lower maintenance and repair costs.
Besides the economic aspects, safety reasons have meant that stainless steel grades used in fixing
engineering have gained tremendously in importance, particularly in complex applications such as
road tunnels.(3), (4), (5), (6), (7), (8) and (9) Stainless steel grades are used more frequently for conventional
architectural applications such as claddings, the reason being environmental impact factors,
improved safety standards and aesthetics.
7.1.2 STAINLESS STEEL THAT RUSTS
Corrosion resistance increasing
According to the new approval, only materials with

the highest corrosion resistance class IVY must be
used in atmospheres containing chloride and in
inaccessible, corrosion-prone places such as road
tunnels. TABLE 14 on page 40 shows an extract from
the General Building Supervision Approval DIET Z30.3-6 with the different stainless steel grades used
in construction engineering shown in relation to the
corrosion resistance classes. FIGURE 17 at right also
attempts to classify some of the listed steel grades
used in fixing elements in terms of their corrosion
resistance and to compare their strengths in their
solution-annealed condition.
See page 40.
Pure austenite
+ Mn + N - Mo - Ni
1.4529
Pure austenite
+ Mo + N
1.4539
+ Cr + Ni - N
- Ni - Mo
1.4439
Pure austenite
1.4462
Duplex with 50% austenite + 50% ferrite

Austenite with ferrite traces
+ Mo + N
1.4571
Austenite with maximum 7% ferrite
+ Ti
1.4401

+ Mo
1.4571
+ Ti
1.4301
0

100
200
300
0.2% elongation limit

in Nm/mm2 (minimum
value according to norm)
400
500
STRENGTH
Using a specific application and the results of long term exposure test and laboratory trials, this
section attempts to show how safe and sustained material concepts can be used for fixing
engineering in tunnel construction.
39
(8)-(12)
1.4565
CLASS
Quite a number of results from exposure tests

and special tests in pollutant gas atmospheres
allow conclusions to be ascertained in terms of
the long term life of various stainless steel
grades in road tunnels. These findings have
also led to a new differentiation in stainless steel
grades being introduced in the General Building
Supervision Approval DIET Z-30.3-6 on 3 December
2003 (Germany).
CLASS 1 CLASS 1 CLASS 1 CLASS 1
When exposed to extreme ambient conditions,however, it has been seen that even stainless or
non rusting steel grades can rust or corrode, with a relatively minor tension existing, for instance,
as intrinsic tension in the actual material, normally sufficient to trigger tension crack corrosion.
Ambient atmospheres with corrosion-inducing pollutant are often seen in industrial surroundings, in
the chemical industry, in multi-storey car parks, indoor swimming pools, in chimneys and in tunnels.
In indoor swimming pools, it is mainly the high chloride concentration,(10) and (11) while other specialised
applications are also exposed to high pollutant concentrations which constitute a substantial
corrosion hazard by forming some highly aggressive condensate.(12) Selecting the right kind of
material is often difficult for aggressive atmospheres such as these, with planners having to rely on
field tests and/or on adapted special tests to determine the expected corrosion behaviour of the
various materials in these applications.
FIGURE 17:
ALLOY STRUCTURE AND STRENGTH
PROPERTIES OF STAINLESS STEELS
7. FIXINGS
Source: Fischer fixing systems
TABLE 14: EXTRACT FROM GENERAL BUILDING SUPERVISION APPROVAL DIBT Z-30.3-6
NOTES:
Under DIN EN 10088-1 or
SEW 400.
(2)
A = Austenite;
F = Ferrite;
FA = Ferrite-austenite.
(3)
Only for blank metallic surfaces. In the event of potential contact corrosion, the
less noble metal is at risk.
(4)
(5)
Constructions are classified

as inaccessible if their condition cannot be checked or
can be checked only under
aggravated conditions and
which, if required, can be
refurbished only at great
effort and expense.
These materials have a high
resistance to tension corrosion. The materials 1.4565,
1.4529 and 1.4547 also
have a higher resistance to
local corrosion manifestations (pitting and/or fissure
corro-sion). For components
in indoor swimming pool
atmospheres without regular
cleaning, only the materials
1.4564, 1.4529 and 1.4547
are suitable. In areas with
water with a CI content
<250mg/litre (drinking water),
material 1.4539 is also
permitted.
REFERENCES:
(8)
Arnold, N., Gmpel, P. &

Heitz, T. (1998). Chloridinduzierte Korrosion von nicht
rostenden Sthlen in Schwimmbadhallen-Atmosphren. Teil 2: Einfluss von
Hypochloriten Materials and
Corrosion 49, S. 482/488.
(9)
Arnold, N., Gmpel, P. &

rostenden Sthlen in Schwimmbadhallen-Atmosphren. Teil 3: Einfluss einer
realen Schwimmhallenatmosphre; Materials and
Corrosion 50, S. 140/145.
(10)
(11)
(12)
Pscheidl, P. (1996). Korrosionsverhalten hochlegierter

nicht rostender Sthle in
ausgesuchten Umweltkompartimenten. Diplomarbeit
an der Fachlochschule
Konstanz, Konstanz.
rostenden Sthlen in Schwimmhallen-Atmosphren.
Diplomarbeit
an
der
Fachlochschule Konstanz,
Konstanz.
Htterer, H. (1994). Schadgaskorrosion von Werkstoffen in der Befestigungstechnik. Diplomarbeit an der
Fachlochschule Konstanz,
Konstanz.
Corrosion
Steel grade(1)
W-No.
Microstructure(2)
X2CrNi12
1.4003
X6Cr17
1.4016
X5CrNi18-10
1.4301
X2CrNi18-9
1.4307
X3CrNiCu18-9-4
1.4567
X6CrNiTi18-10
1.4541
Brief designation
X2CrNiN18-7
1.4318
X5CrNiMo17-12-2
1.4401
X2CrNiMo17-12-2
1.4404
X3CrNiCuMo17-11-3-2
1.4578
X6CrNiMoTi17-12-2
1.4571
X2CrNiMoN17-13-5
1.4439
X2CrNiMoN22-5-3
1.4462
FA
X1NiCrMoCu25-20-5
1.4539
X2CrNiMnMoNbN25-18-5-4
1.4565
X1NiCrMoCuN25-20-7
1.4529
X1CrNiMoCuN20-18-7
1.4547
Resistance
class(3)
Load and typical

application
I/low
Interior constructions with the

exception of humid rooms.
II/moderate
Accessible constructions,
with negligible chloride and
sulphur dioxide content, no
industrial atmosphere.
III/medium
Constructions with moderate

chloride and sulphur dioxide
exposure and inaccessible
constructions.(4)
IV/strong
High corrosion load(5) caused

by chloride and/or chloride
or sulphur dioxide and high
air humidity, rising
concentrations of pollutants.
7.1.3 DUST ATTACKS ANCHORS

One of the most significant factors is the dust load acting on the fixing element. Tunnel dust not only
includes soot, but also mineral substances, abrasive particles from tyres and wearing parts as well
as several percentages per weight of water-soluble chloride originating from salt spraying the road
during winter months. This salt, bound in dirt and snow residues, is normally carried into the tunnel
by the vehicles and more or less spreads evenly throughout the tunnel. Empirical data shows that
the chloride load in the dust can be lowered at certain points by regular cleaning, but that the
chloride content is still 1% by weight at these points.
7.1.4 SALT FILM ON MATERIALS
The salt crystals carried in dust form a highly concentrated salt film on the surface of the materials,
particularly under the condensate conditions usually found in tunnel atmospheres. Some data is
available on the impact of unadulterated salt load on stainless steel grades, with laboratory tests
examining the effect of dehydrated salt droplets from various chloride solutions, such as MgCl2,
NaClO and Ca(ClO)2 solutions on various stainless steel grades exposed to mechanical tension
when stored in humid air at a temperature of 40C.
TYPE OF CORROSION
(1)
1st and 2nd test series
Tension crack
corrosion
3rd and 4th test series
Localised
corrosion
(pitting)
Area
corrosion
1.4401
1.4462
1.4539
1.4529
1.4565
MATERIAL NO.
FIGURE 18: TYPES OF CORROSION FOUND DURING EXPOSURE OF

STAINLESS STEEL SPECIMENS WITH SALT DROPLETS IN HUMID AIR
FIGURE 18 at left
shows clearly that the
resistance of the
materials rises in line
with the corrosion class
(FIGURE 17 on page 39).
These tests also
showed a good
correlation between the
corrosion behaviour
and the so-called
effective sum which
can be calculated from
the composition of
the alloy using the
following equatio:
W = %Cr + 3.3
%Mo + 30 %N
1st and 2nd test series in 70% relative humidity,

3rd and 4th test series in 35% relative humidity
40
A purely qualitative classification of the corrosive intensity judged by the appearance of the
specimen after long-term exposure clearly shows the correlation between the effective sum of the
steel and its corrosion resistance (FIGURE 19, below right). The relative classification between no
corrosion attack (stage 0 in FIGURE 19) and a very strong corrosion attack (stage 12 in FIGURE 19)
demonstrates that the attack with higher effective sum is a great deal lower.
FIGURE 19:
RELATIVE CLASSIFICATION OF THE CORROSION
AFTER LONG-TERM EXPOSURE AS FACTOR OF
THE EFFECTIVE SUM OF THE STEEL GRADES
CORROSION CLASS
(relative unit)
14
50%
45%
12
40%
10
35%
30%
25%
20%
15%
10%
5%
0%
1.4401
1.4462
1.4539
1.4529
1.4565
MATERIAL NO.
2nd test series (70% relative humidity)
4th test series (35% relative humidity)
Effective sum of the steel composition
7.1.5 CORROSION CAUSED BY POLLUTANT GASES
The resistance of materials has been tested in

numerous test series using climate and pollutant gas
test chambers. The pollutant gas concentrations used in these tests are shown in TABLE 15 on
page 42. The climate cycles used in the tests are shown in FIGURE 26 below.
FIGURE 20:
TENSION CRACK
CORROSION IN 1.4401
FIGURE 26: CLIMATE CYCLES OF THE 4TH TEST SERIES

50C
TEMPERATURE
Relative humidity
100%
40C
80%
30C
60%
20C
40%
10C
0C
0 hour
TIME
1 hour
2 hours
3 hours
4 hours
5 hours
Relative humidity
Temperature
20%
6 hours
Following the tests, which were

carried out both with round
specimens and in pre-stressed
bow type specimens, the
surfaces of the material were
checked visually and the degree
of corrosion was classified
according to a guideline series.
FIGURES 27 A - D on page 42
show the classification of the
corrosion of the various
stainless steel grades. It also
demonstrated that it is only the
high-alloy steels 1.4529 and
1.4565 which show a high
resistance to tunnel atmospheric
pollution.
To quantify the corrosion, the round material specimens were cleaned in a light pickling solution
which removed the loose, porous and flaky corrosion layers, such that only the attack of the
localised corrosion (pitting) remained. The depth of corrosion was then measured using a
measuring microscope, with the median value of 5 different pitting points (if available) in the round
specimens determined. The results of these measurements are shown in FIGURE 27 for all materials
tested.
The high-alloy materials 1.4529 and 1.4565 also remain corrosion-immune in this test and therefore
showed the highest resistance under exposure to pollutant gases in road tunnels. All of the results
correlate well in total with the results of the exposure tests in road tunnels and with empirical data.
Among the alpine tunnels, the strongest corrosive effects were found in the Mont Blanc and the
Gotthard tunnel where the resistance of the materials 1.4529 and 1.4565 was also demonstrated,
whereas the other high-alloy steels such as duplex 1.4462 showed localised corrosion.
FIGURE 21:
SELECTIVE CORROSION
IN 1.4462
FIGURE 22:
TENSION CRACK
CORROSION IN 1.4439
FIGURE 23:
LOCALISED PITTING
CORROSION IN 1.4539
FIGURE 24:
BEGINNING TENSION CRACK
CORROSION IN 1.4539
FIGURE 25:
MODERATE TOUGH
CORROSION IN 1.4529
41
EFFECTIVE SUM
The use in tunnel atmospheres with deposits

containing chloride requires the appropriate
resistance to the highly dangerous localised corrosion
(pitting) and tension crack corrosion. For the materials
1.4462 and 1.4539, classed in the General Building
Supervision Approval DIBT Z-30.3-6 as conditionally
corrosion resistant in the highest corrosion resistance
class IV, these requirements are not met, with the
effect that these materials should not be used at
points exposed to the highest loads. FIGURES 20-25 at
right below show examples of the specimens of the
various stainless steel grades prepared with dried salt
droplets and stored in humid air at longer periods. Of
the steel grades examined, it should be noted that it
is only the materials with the highest effective sum,
i.e., steel grade 1.4529 with 6% molybdenum and
steel with high nitrogen content 1.4565, that show an
adequate corrosion resistance to exposure of this
description. The situation can be exacerbated further
by porous particles of soot because this will enlarge
the cathode surface.
7. FIXINGS
Source: Fischer fixing systems / Continued from page 41
TABLE 15: EXTREME VALUES OF ATMOSPHERIC COMPONENTS IN ST. GOTTHARD TUNNEL

AND IN POLLUTANT CONCENTRATIONS IN TEST SERIES
Atmospheric components
S2
(ppm) max.
NO
(ppm) max.
NO
(ppm) max.
St. Gotthard Tunnel
>20
14
1st test series
45
30
2nd test series
45
3rd test series
22
4th test series
43
FIGURE 27 A
7
FIGURE 27:
BEHAVIOUR
OF STAINLESS 6
STEELS IN THE
POLLUTANT
5
GAS TEST
H2S
(ppm) max.
Chloride
content
2.0
Max. 2.5%
13
5%
30
13
5%
14
2.3
2.5%
30
11
Inside
Outside
B
3
1
1.4003
FIGURE 27 B
See indicative figures

at right of the page
1.4301
1.4301
1.4571
1.4439
1.4565
1.4529
MATERIALS OF THE BOW-TYPE SPECIMENS
SUMMARY
In principle, stainless steel should be used for fixing elements only if it has the appropriate corrosion
resistance. The long service life of such stainless steel grades can help prevent major financial
losses caused by corrosion. The corrosion resistance of stainless steels is a typical and systeminherent property and depends to a large extent on the various system parameters. The system
parameters coming to bear in the corrosive effects acting in road tunnels (FIGURE 28 below) are
subject to wide-ranging scatter, which is certainly one reason why the empirical values of the
various tunnels differ so widely. Consequently, the views regarding the use of material grades and
qualities in road tunnels differ equally widely. Unlike in structural steel, the process of corrosion in
stainless steel does not show as rusting off with the resulting reduction in the cross section, but
by way of a selective corrosion occurring after a short period of time. Stainless steel for fixing
elements in road tunnels should therefore always be selected in compliance with the regulations in
the German General Building Supervision Approval Z-30.3-6 which will ensure that the
requirements involving the use of these high grade materials are met.
DEPTH OF CORROSION
FIGURE 28: MAXIMUM DEPTH OF PITTING CORROSION IN m OF THE ROUND MATERIAL SPECIMENS
IN THE LABORATORY POLLUTANT GAS TEST
70
FIGURE 27 C
60
50
40
30
20
10
0
MATERIAL NO.
1.4003
1.4104
1.4305
1.4541
1.4301
1.4571
1.4404
1.4404
1.4435
1.4462
1.4462
1.4565
1.4529
42
FIGURE 27 D
7.2 CRACK WIDTH IN CONCRETE

In accordance with many international standards (e.g. BS, DIN, Euro codes), the maximum acceptable
crack width in reinforced concrete is limited to wk=0.3mm under semi permanent loadbearing
conditions. If structures are subject to exceptional loading, e.g. seismic, then there is a possibility that
wider cracks could occur. Recent analysis shows that the cracks in reinforced concrete structures could
be as wide as 1.5mm after being subjected to earthquakes to the maximum design load.
Anchors might be situated near to, or even within cracks in the concrete, and as worst case, could even
be positioned at the intersection point of two cracks. Tests have confirmed that in the event of concrete
cracking, there is a high probability that at some point, these cracks would radiate into contact with the
fixings. This is particularly true of expansion bolts because there are local tensile stresses in the area
surrounding the fastening due to the expansion forces of the bolts.
The type of bolt to be used should be chosen with due consideration of the possibility of the concrete
cracking, and an appropriate type of fixing utilised.
7.3 COMPRESSIVE STRENGTH OF CONCRETE

If a tunnel is constructed using high strength concrete, the nominal compressive strength of the
concrete would be approximately 60N/mm2. However, it is often the case that the compressive strength
is far higher than this, at around 90 to 100N/mm2.
Anchors in which the fastening is by means of torque control or displacement controlled friction locking
are not suitable for use in this type of high strength concrete. Torque controlled anchors are not capable
of creating the deformation in the concrete which is required for the bolt head to expand into. Thus the
load would be held by friction of the bolt within the drilled aperture only. A drilled hole tends to be
smooth, the friction between the anchor and the hole is minimal and the loading which can successfully
be applied is unacceptably low.
For fixing into high strength concrete, undercut anchors should be used, as these do not rely on the
compressive strength of the concrete, nor the smoothness of any drilled hole.
With the research that has taken place over the last ten years or so, both by fixing manufacturers and
independent research bodies, it has been determined that the optimum material for fixings in tunnels
should consist of stainless steel of grade 1.4529. This material is resistant to all types of corrosion,
surface pitting and corrosion induced stress cracking.
43
8. METHODS OF APPLYING BOARD PROTECTION

8.1 AS LOST SHUTTERING
FIGURE 29:
BOARDS LAID WITH THE
SMOOTH FACE DOWN
USING LOST FORMWORK
(as illustrated below)
8.1.1 PROMATECT-H CEILING SYSTEM, LOST FORMWORK METHOD

Essentially this system consists of the following installation steps:
A) THE BOARDS ARE LAID ON THE LOADBEARING FORMWORK
The PROMATECT-H boards will be laid with the smooth face down (FIGURE 29) as this will provide
the fair faced finish after completion of the tunnel.
It is very important to align the first row of boards alongside a straight edge reference marker,
which is screwed to the formwork (FIGURE 30). The rest of the boards are laid next to each other,
with butt joints, utilising the previous row as the next starting point or datum. No special treatment
on the joints is required (FIGURE 31).
The dimensions of each tunnel section in the ceiling area are approximately 16m x 25m on a typical
Dutch cut and cover or immersed tunnel. If the dimensional and squareness tolerances on the
PROMATECT boards are too large, gaps occur while laying the boards as the tolerances tend to
accumulate and add up. In order to avoid this effect, PROMATECT boards are cut to tight
tolerances in the factory, such that gaps in between the boards will be minimised.
Where the tunnel is designed with sloped sections, the
so-called haunches, the edges of the boards are simply
cut at an angle and installed butt jointed (FIGURE 32).
In order to minimise tolerances by cutting on the job site,
there are two options. One is to lay the boards on the
formwork, draw the cutting line (for example, at the end
of a section) and cut all boards in one go alongside a
datum on the formwork. The second option is to pre-cut
the boards in a dedicated on site workshop.
FIGURE 30:
FIXED STRIP OF PLYWOOD SERVES AS A
DATUM FOR THE FIRST ROW OF BOARDS
The boards can either be installed using staggered joints

or straight joints. Experience from contractors indicates
that staggered joints result is less gaps between the
boards This method allows compensation for certain
tolerances, whereas the straight joint method does not
allow for much compensation.
FIGURE 32:
BOARDS CUT TO AN ANGLE AT THE HAUNCHES
FIGURE 31:
BOARD JOINTS ARE SIMPLY ABUTTED,
NO SPECIAL TREATMENT REQUIRED
44
The locations of the screws are marked on the boards, using a template and a spray can of
paint (FIGURE 33).
FIGURE 33:
PRE-MARKED SCREW
LOCATIONS ON THE BOARDS
B) THE FIRST LAYER OF REINFORCEMENT IS INSTALLED

Prior to the installation of the stainless steel screws, the first layer of reinforcement is installed on
the stools (spacer blocks creating concrete cover thickness). In this way the screws are always
protected from foot traffic (FIGURE 34).
C) STAINLESS STEEL SCREWS ARE PARTLY INSERTED IN THE PROMATECT-H BOARDS
The 50mm long screws are inserted to a depth of 20mm,
through the openings in the reinforcement; the remaining
30mm projects out of the board creating the anchorage to
the concrete after it has been poured (FIGURE 35).
FIGURE 34:
FIRST LAYER ON
REINFORCEMENT INSTALLED,
PRIOR TO SCREW INSERTION
Several methods are available on the market to ease the

installation of the stainless steel screws (FIGURES 36 & 37):
Battery-powered screwdrivers can be equipped with
depth guiding devices to ensure the correct depth of
20mm.
Screws can be supplied on a plastic strip, which is
fed into the screwing machine, increasing installation
speed.
FIGURE 36: SCREW INSERTION
Battery-driven drilling machines can be equipped with

a tube through which the screw can be dropped down
onto the surface of the board. The installer remains
standing upright while installing the screws.
A combination of the above features can be made and
should first be discussed with a local power tool supplier.
The screws for the lost formwork system are 5mm x 50mm
chipboard screws, with a countersunk (CSK) head type and
a Pozidrive-2 connection.
1250mm
50mm 288mm
287mm
287mm
288mm 50mm
FIGURE 37:
SCREW INSTALLED AND PROTECTED
BY REINFORCEMENT
343mm 50mm
343mm
The number of screws should be in the region of 12 screws per m2. In order to ensure the same fire
performance in practice, the exact same set-up as during the fire test should always be followed,
including fixing materials.
343mm
The screw pattern of a full size board of 2500mm x 1250mm, as detailed in FIGURE 38 for example,
has an average of 12.8 screws per m2.
342mm
The majority of the boards are laid on the formwork in their standard full size dimension. Some
panels will have to be cut to size in order to cover the whole surface of the formwork and to connect
to the walls and construction joints. Where cut to size panels are used, the following criteria should
be followed:
The minimum quantity of screws should be 12 screws per m2.
343mm
The distance from the edge of the boards should be a minimum of 50mm.
343mm
The screws for cut pieces should be evenly distributed over the surface of the panel. In other
words, the spacing distances in X and Y direction should be optimised and be as close
as possible.
50mm 343mm
For example, a screw pattern on a cut to size panel of 1675mm x 1090mm in which all criteria are
followed, the surface area of the panel is 1.83m2 which means that the minimum member of screws
should be 22.
FIGURE 35: EXAMPLE OF

SCREW PATTERN OF LOST FORMWORK
45
2500mm
The design of a screw has major influence on mechanical

performance and fire performance, in combination with
PROMATECT boards. Among others, properties like shaftdiameter versus thread-diameter, sharpness of the tip and
distance of the winding (thread) will influence the stickability
of the screws and the performance of the final system.
FIGURE 35:
PRINCIPLES OF
SCREW INSERTION
D) WHEN THE CONCRETE IS POURED

Before the concrete is poured, the PROMATECT boards should be hosed down to remove
accumulated site debris and to moisten the boards to minimise water absorption from the concrete
mixture. Excessive water should also be removed from the surface of the boards prior to the
pouring of the concrete (FIGURE 39).
During vibration of the concrete, the machinery being used is best kept away from the surface of
the PROMATECT boards.
FIGURE 39:
THE POURING OF CONCRETE
E) AFTER THE CONCRETE IS SUFFICIENTLY CURED, FORMWORK IS EXTRACTED

The following advantages of PROMATECT as lost formwork system have been reported by
contractors over the years:
FORMWORK SAVINGS
The shuttering material only has to have load-bearing properties. There is no need to apply
phenol coated plywood boards as PROMATECT boards will be laid on top of the formwork.
The formwork elements just have to be installed properly (i.e. level and flushed).
As concrete will not be in direct contact with the formwork, there is no need for demoulding
oil (i.e. no slippery surfaces). The plywood sheets will remain clean and can be re-used.
Formwork can be installed at a distance up to 90mm from the side walls.
FIGURE 40:
EXTRACTED FORMWORK
The PROMATECT-H boards can span the 90mm distance, depending on the expected load.
At this location, care should be taken with the vibrating action during pouring of the concrete.
Extraction of the formwork is much easier as it will not get jammed between the walls.
As the PROMATECT forms a barrier, there is no adhesion between the concrete and the
formwork, it is easy to extract the formwork and it remains clean (FIGURE 40 at left).
EASE OF INSTALLATION
Joints between boards only have to be butt jointed. No special treatment (e.g. fillers or
mastic) is necessary from a fire performance point of view. The cement water will not run
through the joints. Where gaps of more than 1mm occur, mastic can be used to seal the gap
in order to prevent water from the cement leaking through the gap, causing unsightly stains.
Vertical wall panels can also be installed using the lost formwork system.
Curves in the tunnel can easily be dealt with by cutting the boards on the formwork at an
angle to accommodate the curve.
FIGURE 41:
OPENING OF MANHOLES
Openings in the PROMATECT lining for manholes (FIGURE 41 at left) and end-walls can easily
be made by installing phenol coated formwork instead. After striking the formwork, standard
size PROMATECT boards will close the opening by post fixing the board into the opening.
See SECTION 8.2 on pages 47 to 53.
PROMATECT-H boards provide a heavy duty floor surface. The abrasion resistance is such
that the surface can withstand the exposure to people walking and working on top of it, even
in wet conditions. Also, the weight of bundles of reinforcement steel and pallets of materials
will not cause any damage to the boards. This assumes the PROMATECT boards are
adequately supported by the formwork.
Rapid installation method. Installation rates for the PROMATECT boards of 150m2 per man
per day have been reported on European tunnel projects.
The installation of the system does not interfere with other construction activities.
Extensive Promat expertise is available with this system in immersed and cut and cover tunnels.
FIGURE 42:
ANCHOR SYSTEM
FOR JET-FANS
TUNNEL SERVICES AND SPECIAL SHAPES

Anchor systems for services (e.g. jet-fans, FIGURE 42 at left) can be fixed onto the upper
surface of the boards, prior to casting the concrete.
Services, pipes, tubes etc. can be included within the depth of the concrete.
After the formwork is extracted, services can be installed onto the PROMATECT lining from
below. Anchors can be installed through the boards into the concrete, thus providing a
continuous fire protective layer.
Special shapes in the concrete structure can easily be accommodated, e.g. beams.
OTHER BENEFITS
As the PROMATECT boards are installed in the very early stages of the construction of the
tunnel, fire protection during the construction phase is provided.
The PROMATECT lost formwork system provides a flush finish ceiling.
No obstacles such as anchor heads on the soffit of the tunnel.
46
8.1.2 PROMATECT-T CEILING SYSTEM, LOST FORMWORK METHOD

The lost formwork method can also be applied, utilising PROMATECT -T boards.
The system and the installation method are similar to the PROMATECT-H system as described in
SECTION 8.1.1, with the following additional advantages:
The very high thermal performance of PROMATECT-T boards is unmatched. Therefore the
thickness of the fire protective layer can be reduced and the interface temperature on the
concrete surface will be even lower, thus providing a higher level of structural safety.
The PROMATECT-T boards can be supplied with red coloured X-marks at the location of
the insertion of the screws, on the reverse side of the board. For standard board dimensions
of 2500mm x 1200mm, the boards can be supplied with 5 rows of eight X-marks. The
tolerance on the location of the X-mark is 10 mm.
2
1
Half size boards of 1250mm x 1200mm can be supplied for the ease of handling.
PROMATECT-T boards are cut to tight tolerances in the factory, suitable for application in
a lost formwork system.
The PROMATECT-T lost formwork system is the most competitive lost formwork system in
the market capable of withstanding the Rijkswaterstaat (RWS) fire exposure.
PROMATECT-H or PROMATECT-T
boards fixed to the underside of the
concrete
Expansion bolts or alternative fixings of a

length and at appropriate centres. It
should be noted that bolts are used in
conjunction with washers to prevent
penetration of the bolt head into the board.
Steel reinforcement
Concrete slab with minimum 25mm cover

to the reinforcement. Cladding thickness
concrete grade and type dependant.
8.2 POST CONSTRUCTION INSTALLATION TO CONCRETE

In many instances, the construction method used to build a tunnel prohibits the installation of the
PROMATECT-H using the lost shuttering method. It is also the case that many older tunnels may simply
require upgrading. To this end, Promat has also developed and tested systems for the protection of
concrete where the fire protection boards are applied after the structure has been completed.
An example for such a post construction method is the Zeeburg Tunnel in the Netherlands.
FIGURE 44: POST CONSTRUCTION METHOD IN
ZEEBURG TUNNEL, NETHERLANDS
47
FIGURE 43 (above) :
POST INSTALLATION OF
SINGLE LAYER PROMATECT BOARDS
8.2.1 PROMATECT-H OR PROMATECT-T FOR POST CLADDING

PROMATECT-H boards are available in dimensions up to 3000mm or 2500mm x 1250mm and
PROMATECT-T boards are available in dimensions up to 2500mm x 1200mm.
The PROMATECT should be installed with the fair face of the board looking down into the
tunnel. The boards should be placed into position and carefully supported while the holes for
the bolts are being drilled and the bolts inserted. Although PROMATECT boards are relatively
small in size, the thicker boards are of a reasonably substantial weight, e.g. 1200mm x 1200mm x
27mm PROMATECT-H weighs approximately 36kg, so installation should be considered as a two
person operation.
PROMATECT-T boards can be curved at site, depending on the diameter of the tunnel and the
thickness of the boards. Where a thicker board is required to provide the specific fire resistance
level, it is possible to install the PROMATECT-T boards in multiple layers of thinner panels in order
to make up the required thickness while still allowing the panels to be curved at site.
An example for curved post cladding method is the Clyde Tunnel (FIGURE 45 at left) in Glasgow,
Scotland, where PROMATECT-T boards were used to line the cast iron tunnel sections.
8.2.2 FIXINGS
It is likely that the concrete to which the PROMATECT boards is being fixed would not be
completely flat. Therefore care needs be taken when fixing the boards to ensure the removal of any
large nibs of concrete. In addition, the bolts fixing the boards should be carefully tightened to avoid
over turning and cracking of the boards where positioned on uneven surfaces.
Bolts should be installed a minimum of 100mm from edges of the boards, and should not be
located directly in the corners of the boards. Bolts should be offset to avoid cracking or breakage
at the corners.
FIGURE 45:
APPLICATION OF CURVED
PROMATECT-T TO CAST IRON
TUNNEL LINING SECTIONS
IN CLYDE TUNNEL,
GLASGOW, SCOTLAND
Bolts used in the installation of PROMATECT boards should be used in conjunction with washers
of a minimum of 20mm diameter, or should have their own integral washer, to prevent the heads of
the bolts being driven into the surface of the boards. Note that washers should be manufactured of
the same material type as the bolts to ensure that corrosion does not occur.
Care should be taken when drilling holes into the concrete to avoid the positions of the
reinforcement within the concrete. The PROMATECT-H boards should, as far as is possible, be
properly supported when drilling takes place to ensure the rear face of the boards do not blow at
the exit point of the drill bit.
Minimum requirements for anchors used to secure PROMATECT boards:
M6 in diameter.
Made of stainless steel of 316 grade or higher.
Appropriate length to secure the panel thickness (see SECTION 7 on pages 39-43).
Minimum 40mm anchor depth penetration into the concrete.
Expansion action of the anchorage shall be within the concrete and not within the
PROMATECT panel.
Supplied with a nut and washer head to facilitate removal of the PROMATECT panels
where required.
Suitable for use in tension zone of concrete (cracked concrete).
Suitable for use where anchors will be subject to positive and negative pressure fluctuation
(dynamic loads).
For corrosion resistance of fixings, please refer to SECTION 7.
8.2.3 STEEL FRAMING
Tests have been carried out on systems for both horizontal and vertical applications where
the PROMATECT-H or PROMATECT-T boards have been fixed to a steel sub-frame. The
type of steel used as the framing is of course dependent upon environmental conditions of the
tunnel but would generally be of a grade of stainless steel consistent with the corrosion
resistance requirements.
The steel frame for horizontal applications is generally designed with a number of considerations in
mind. For example, is the frame fixed directly to the concrete structure or is it a free span across
the width of the tunnel (see SECTION 8.2.6 on pages 51 and 52). The frame could consist of either
zed sections or top hat (omega) sections positioned at nominal 600mm or 625mm centres fixed
directly onto the concrete soffit or it could consist of steel channels or hollow sections if it is to span
across the tunnel. In either case, the fixing of the steel frame would be subject to the exact same
considerations as for the direct fixing of PROMATECT board (see above SECTIONS 7 & 8.2.2).
48
Fixing type, centres and depth into the substrate depends upon the type of framing system and is
subject to the fire performance requirement and substantiation by fire test reports. Please consult
Promat for details pertinent to any specific installation.
It is not possible to provide a definitive statement on the types of steel framing for vertical wall
systems, as these are again dependent on a number of factors, e.g. the fire performance
requirement of the system etc. Consideration must be paid to the same factors affecting horizontal
applications, with both orientations required to resist the effects of wind loading and suction
induced by the passage of traffic.
The framing systems employed tend to be designed on a project by project basis because the
section size of the framing is determined by the effects of suction forces, as well as the height of
the construction, and the need for protection to any services that may be located behind the lining
system. Thus the dimensions and shape of the steel supporting section are determined by the
section modulus required to be capable of resisting the compressive loads, bending moments and
other forces which may be imposed on the wall lining.
8.2.4 INSTALLATION OF PROMATECT-H OR PROMATECT-T BOARDS
A) PREPARATION OF BOARDS
Wherever possible, PROMATECT-H or PROMATECT-T boards should be processed and made
ready for installation when delivered to installation site. The preparation works should be carried out
in a suitably equipped workshop either at an off site location or, if the conditions permit, at an on
site location. However, provision for remedial work should be made available at the installation site
should there be necessity to make changes to dimensions and edges.
FIGURE 46:
PANEL LIFTING HOIST
The board preparation works include the following:

Cutting of PROMATECT-H panels to size according to the requirements of the ceiling plan.
Pre-drilling of holes to make PROMATECT-H panels ready for securing of impact anchors.
The position for holes for various PROMATECT panels are predetermined according to the
anchor layout plan. Suitably prepared templates must be used to drill the anchor layout on
each PROMATECT-H panel.
B) INSTALLATION OF BOARDS
With the smooth face of the PROMATECT panels facing down, the panels are held in positions
flat against the substrate with suitable clamping and lifting equipment, e.g. a panel lifting hoist
(FIGURE 46 at right).
Drill into the concrete to the required anchor depth. Whenever required, a rebar detector should be
employed to ensure that hitting of the reinforcing bars due to discrepancies in concrete cover
is avoided. Anchor positions should be adjusted to accommodate this situation. However, the
required panel area to anchor ratio should be maintained at all times.
Insert impact anchors into the pre-drilled holes, and knock the anchor into position until the
washers are in tight contact against the PROMATECT panel surface. Visually inspect that the
anchors are tight and secure. Any dislodged anchors must be replaced. Care shall be taken not to
over drive the anchor and damage the PROMATECT panel. Place the next PROMATECT panel
tightly abutting the installed panel and repeat the process. Repeat installation of panels outwards
from the inner tunnel wall and towards the outer tunnel wall. See FIGURE 47 at right for examples of
a machine drilling the concrete.
Care must be exercised to ensure that the butt joints between panels are as close as possible.
Visually judged gaps of 1mm to 3mm are acceptable. Gaps shall not exceed 3mm. Where gaps
cannot be kept within the maximum due to site discrepancies, PROMATECT-T tunnel joint
compound should be used where necessary to make good any minor joint misalignment.
If a situation arises where it is impractical to use pre-drilled PROMATECT panels as templates, the
template used for off site drilling can be used to facilitate the simultaneous drilling of both the
PROMATECT panels and the anchor positions. Thereafter secure the impact anchors as described
above. Pre-cut or cut on site panels shall be prepared to suit site conditions for panels along the
junction with tunnel walls.
49
FIGURE 47 (above and this pictures) :

DRILLING OF CONCRETE

8.2.5. SUSPENDED CEILINGS, PROTECTIVE MEMBRANE
Many tunnels, especially older city tunnel, were built using a cut and cover method and constructed
by means of steel and/or concrete roof beams with a concrete slab or a composite steel/concrete
slab on top of the beams. In many tunnels, the space between the beams is utilised to install pipes,
cable trays and other services.
In refurbishing such a tunnel, the protective membrane system is both technically and commercially
the most feasible option.
A) SYSTEM COMPONENTS
The protective membrane system consists out of a steel frame which is suspended from the load
bearing structure or, depending on the span of the ceiling, can be supported along the walls only
(FIGURE 48 below). The steel frame should be designed such that it can cope with:
The dynamic load cycles coming from passing traffic.
The additional weight of the PROMATECT boards, also taking into account the potential
additional weight of water which may be absorbed into the boards.
The elevated temperatures in case of fire and still retain its function.
FIGURE 48:
SUSPENDED CEILINGS,
PROTECTIVE MEMBRANE
The designer of the suspended steel frame has two options for the horizontal
load bearing members:
1) The use of C, Z or omega profiles;
2) The use of a trapezoidal steel decking.
Typically the design of such a suspended steel frame is conducted by a local
structural engineer. The PROMATECT boards are screwed from below to
the suspended steel frame, either the profiles or the trapezoidal steel sheets.
Fire tests have shown that when exposed to an RWS fire curve, for example,
the temperature of the steel frame can still reach some 300C. At this
temperature the steel frame in the fire test maintained its mechanical stability.
Due to this elevated temperature, the thermal expansion of the steel
members could potentially introduce gaps between boards possibly causing
thermal leaks.
In order to address this issue, protective membrane systems are equipped
with cover strips at the joints. These cover strips can be installed either
behind or in front of the boards. If the latter option is chosen, the advantage
is that no intermediate strips are required and the installation rate can be
increased.
B) THERMAL DATA
Fire tests have been conducted on both systems as described above, i.e.
a frame with profiles and a frame with the trapezoidal steel decking. In both
situations the PROMATECT-T boards have been attached from below,
using cover strips at the joints.
During these fire tests temperature recordings have been taken on the
following locations on and within the system:
1) Reverse side of the PROMATECT board;
2) Reverse side of the trapezoidal steel sheet or on the C-profiles;
3) HEA 350 I-profile, which was supporting the concrete slab, simulating a large steel beam;
4) Lattice girder, which was supporting the concrete slab, simulating a light weight steel
support member;
5) The surface of the concrete;
6) Air temperature in the cavity of the system.
The temperature development on these individual members are available from the Promat Technical
Department. The elevated temperatures on the steel members, as mentioned under 2, 3 and 4 are
of particular interest to the structural engineer designing the suspended steel frame. Based on the
mechanical load, the span, the loading system, the required safety factor and the maximum
temperature, the required steel dimensions can be calculated.
The temperature on the concrete surface, as mentioned in 5 (above), is of interest to address the
reaction of the concrete when exposed to these elevated temperatures. In accordance with the
RWS standard, for example, the maximum allowed temperature on the concrete surface is 380C,
for cast-in-place concrete, which is often applied in such tunnels. If in a particular project the
temperature on the concrete is set at a certain maximum, Promat can advise on the required
material thickness of the PROMATECT-T boards, in order to meet the design criteria.
50
Finally, the air temperature in the cavity of the system can be used to analyse if the maximum
allowable temperature on services is exceeded.
Critical electrical cables for example, are regularly installed behind the protective membrane. Such
cables can be feeding jet fans, emergency lighting and other power operated systems that should
maintain full functionality when exposed to fire. In the design of a protective membrane it should be
determined if the maximum allowable temperature on such services is exceeded.
It should be noted that the maximum failure criteria for the structural members discussed here can
vary widely. As stated above, cast-in-place concrete is perceived to be safe below 380C, whereas
loadbearing structural steel beams are able to withstand elevated temperatures up to 550C,
depending on the mechanical load, the span, the loading system and the required safety factor.
Fire test results have shown that the maximum recorded temperatures on the steel members and
the concrete surface are in line with their respective maximum failure criteria.
Non fire rated cables however can only take some 130-160C. It could therefore be a more
economical option to design the fire resistance of the protective membrane such that the
suspended steel frame, the steel or concrete beams and the concrete slab are sufficiently protected
and to protect the cable trays separately.
8.2.6 SUSPENDED CEILINGS, ESCAPE ROUTES
Typically in circular tunnels the tunnel roof space can be utilised to create an escape route above
the tunnel tube by means of constructing a suspended ceiling
system. The frequent lack of space to provide a means of
egress alongside the tunnel tube means this method is
n
becoming more commonly used in this type of
actio
oute
extr
tunnel.
e
pe r
k
a
n
o
c
o
i
m
s
t
S sec
E ion
duct
sect
The escape door leading to the stairwell
should be fire resistant to prevent fire
spreading into the escape route. The
spread of smoke and toxic gases into
the escape route will also be
prevented. The spread of smoke
and toxic gasses into any escape
route should be prevented. To
achieve this, the escape route
area is pressurised with fresh air,
creating an over pressure to the
surrounding atmosphere.
The area above the road deck
can be used for escape route
purposes only or it can also be
combined
with
a
smoke
extraction duct. In the latter, a fire
resistant wall separates the escape
route area (fresh air) from the smoke
extraction duct. This wall requires fire
resistance because in the event of fire it
will be exposed to fire temperatures
through the hatches in the smoke extraction
plenum system.
Tunnel s
e
ction
Such an escape route ceiling can either be constructed

out of concrete or steel. Regardless of the selected
construction method, the structural integrity of this ceiling during
fire is of paramount importance because it provides the most important means of egress in a fire
emergency. For those instances where the escape route is constructed out of concrete, please refer
to SECTION 8 in this manual outlining concrete protection.
The other option is to construct the ceiling using a steel frame, which would span from wall to wall,
with intermediate hanger rods if mechanically required. The separating wall can also be constructed
such that it functions as a support system. For obvious reasons, supporting structures should be
avoided in the escape route area.
51
FIGURE 49:
PRINCIPLES OF ESCAPE ROUTE
FOR CIRCULAR TUNNELS
The steel frame should be designed so that it can adequately cope with:
The dynamic load cycles coming from passing traffic.
The additional weight of the the PROMATECT boards, taking into account the potential
additional weight of water which may be absorbed into the boards.
The elevated temperatures in case of fire and still retain its load bearing function
Typically the design of such a suspended steel frame is conducted by a local structural engineer.
Apart from its structural integrity in case of fire, an escape route ceiling has an additional thermal
criterion in that the maximum allowable temperature on the non-exposed face of the specimen, i.e.
the temperature on the floor, should not exceed a certain tenability level. The French tunnel fire
safety standard provides guidance to address this. The maximum allowable absolute temperature
on the floor is set at 60C. This is not a temperature rise above ambient but an absolute maximum.
Promat has designed escape route ceiling systems for use in tunnels and have fire tested a number
of different configurations, using PROMATECT boards.
A) SYSTEM COMPONENTS
The escape route ceiling system is constructed using a trapezoidal steel sheet as the load bearing
layer. From below, PROMATECT-T boards are screwed to Z-profiles and are combined with high
density mineral wool thus providing the required thermal insulation of the system. On top of the
trapezoidal steel decking a metal grid is positioned to provide for a flat, unobstructed surface to
walk upon (FIGURE 50 below). The system described above satisfies the thermal requirement of 60C
on the floor surface as mentioned above.
An additional PROMATECT board can be
applied between the trapezoidal steel sheet and
the metal grid to obtain even lower temperatures
on the floor surface.
B) THERMAL DATA
During the fire tests temperature recordings
have been taken at the following locations on
and within the system:
1) Reverse side of the
PROMATECT board;
2) The Z-profiles;
3) The trapezoidal steel sheet;
4) On top of the metal grid
(criterion failure <60C).
The elevated temperatures on the steel
members, as listed under 2 and 3 are of
particular interest to the structural engineer
FIGURE 50: LOADBEARING CEILING CONSTRUCTION
when designing the steel frame. Based on the
mechanical load, the span, the loading system,
the required safety factor and the maximum temperature, the required steel dimensions can
be obtained.
The temperature development on these individual members are available from Promat.
The elevated steel temperatures on the trapezoidal steel sheet and the Z-profiles will cause thermal
expansion of the steel members and could potentially introduce gaps between the boards in case
a single layer of PROMATECT boards is installed. However, for thermal insulation reasons the
escape route ceiling system is equipped with a double layer of PROMATECT boards which are
installed with staggered joints. Therefore the cover strips as discussed in the SECTION 8.2.5 on
page 50 regarding single layer PROMATECT protective membranes are not required on double
layer PROMATECT escape route ceiling systems.
52
8.2.7 APPLICATION OF CURVED AND FRAMED SECTIONS

PROMATECT-T boards do not necessarily
need to be fixed directly to the concrete soffit
of the tunnel. It is feasible for some fire
performance requirements to use steel framing
members. It should be noted that although the
performance of galvanised steel sections is
adequate under fire conditions, the aggressive
environment encountered within tunnels
suggests that the use of stainless steel framing
members is preferable.
FIGURE 51 at far right shows the
PROMATECT-T boards fixed either directly to
the soffit or onto top hat sections. The
dimensions of the steel sections and the centres
of positioning are dependent on a number of
factors, e.g. fire performance, span, thickness of
PROMATECT-T board and type of concrete.
Please consult Promat for further details.
FIGURE 51:
SECTION THROUGH
FLAT CONCRETE SLAB
PROMATECT-T boards can be supplied as

flat sheets and can be curved on site. Care
should be taken to ensure that the thickness of
the PROMATECT-T board is commensurate
with the diameter of the tunnel lining. If the
diameter is too tight, it may be necessary to
install in a double layer of thinner boards rather
than one single board thickness.
FIGURE 52 at right shows the PROMATECT-T
boards fixed either directly to the soffit or onto
top hat sections of a curved section. The
dimensions of the steel sections and the centres
of positioning are dependent on a number of
factors, e.g. fire performance, span, thickness of
Promatect board and type of concrete etc.
Please consult Promat for further details.
53
FIGURE 52:
SECTION THROUGH CURVED CONCRETE SECTIONS

In any tunnel construction, applying a protective material to enhance the fire resistance of the structure
is only part of the story. On its own, this is not going to prevent the loss of life which might occur if there
is a fire in a tunnel. Additional active and passive systems need to be incorporated into the design to
ensure optimum life safety systems. These would include the following:
Enhancing the fire resistance of the structure
Air supply systems
Smoke extract duct systems
The provision of fire and smoke resistant safe havens in long tunnels
Active and passive detection systems
Fire extinguishing systems
The active systems within tunnels should consist of lighting, signal systems, monitoring cameras, fire
and smoke alarms, loudspeakers, antenna systems (for two way radio communication), hydrants, pump
cellars, escape routes, air supply and smoke extracting systems.
This manual is concerned only with systems pertinent to passive fire protection, e.g. air supply and
smoke extract ducts, escape and cross tunnel fire doors, provision of safe havens and systems for the
protection of cables supplying critical services.
9.1 AIR SUPPLY AND SMOKE EXTRACTION SYSTEMS

As has been shown by many case studies into the cause of death resulting from fire in tunnels, the
majority of these casualties are a result of inhalation of smoke particulates.
Smoke can have wide ranging debilitating effects on people:
1) The atmospheres may be hot; temperature near the seat of the fire may exceed 1000C.
Inhalation of hot gases may cause serious burn injuries to the respiratory tract.
2) Toxic and narcotic gases, such as carbon monoxide and hydrogen cyanide, will be present. At
high concentrations, carbon monoxide will cause rapid death; lower concentrations may bring
about a loss of coordination, particularly on exertion, preventing people reaching escape exits.
3) The atmosphere will contain a low concentration of oxygen; this in itself can bring about
unconsciousness and death but normally the effects of toxic gases predominate.
4) There may be many small particles in the atmosphere that restrict vision.
5) The effects of irritants to the upper respiratory tracts and eyes may impede escape.
Studies on the causes of deaths due to fire indicate that carbon monoxide (CO) poisoning accounts for
roughly one-half of total fatalities. The remaining half is accounted for by direct burns, explosive
pressures, and various other toxic gases. Although the analysis of blood cyanide (which would come
from exposure to hydrogen cyanide) in fire victims is sometimes reported in autopsy data, blood
carboxyhemoglobin saturation, resulting from exposure to CO, is often the only fact provided.
in
ht
g
Li
oke
(Smtem)
s
n
s
-fa sy
Jet tract
ex
sure
s enclo
Servicetion system)
etec
ssive d
It is therefore imperative for long

tunnels to include some form of
smoke extraction system in the
design. Due to the very nature of
the hot gases and particulates any
system is required to remove from
the location, a duct or extraction
system will need to be constructed
in such a manner that it too is
resistant to fire.
(Pa
Signal systems
Post cladding boards

(Passive detection system)
Fire extinguisher
(Active detection system)
FIGURE 53: TYPICAL TUNNEL SERVICES SYSTEMS
54
However, it is not a simple matter of installing ventilation or extract fans and assuming these will perform
the necessary services. Significant research (some 98 tests) carried out in the early 1990s in the
Memorial Tunnel, USA provided some valuable data on the performance of ventilation systems. These
included through natural, semi transverse, fully transverse and longitudinal ventilation systems. Similarly,
fire loads ranged through 10, 20, 50 to 100MW in severity. A few sprinkler/deluge systems were also
tested during this programme.
More recently, a series of tests carried out in the new Benelux tunnel in the Netherlands also focused
on the effects of ventilation on smoke layering and sprinklers water dispersion.
FIGURE 54:
EFFECTS OF AIR FLOW
ON CAR FIRE
In tunnels with longitudinal ventilation systems, the ventilation can have a marked effect on the HRR of
the fire. Investigation and experimentation have shown that longitudinal ventilation within a tunnel can
cause different types of fire to behave in very different ways. The HRR of fires in heavy goods vehicles
in particular can be greatly enhanced, even with low rates of ventilation, whereas the HRR of a car under
the exact same conditions could be greatly reduced. There is no simple method of calculating the
complex relationships between ventilation speeds and increases in heat release rates.
Ventilation can also affect the spread of fire along a tunnel. For example, during the Mont Blanc disaster,
fire spread rapidly from the source of the fire to cars situated some 290m away.
As can be seen from FIGURE 54, the effect of the ventilation results in the fire moving horizontally instead
of mainly vertically. As a result of this action, any vehicles positioned down wind of the fire could possibly
catch alight themselves. The top picture shows the effects of a neutral air flow, the lower picture shows
the effects of a ventilation speed of 2m/second.
While the effects of natural and longitudinal ventilation in tunnels has been subject to some
experimentation, the effects on tunnel fires from semi or fully transverse ventilation is at present less
well known.
In tunnels, there are a number of ways for providing the extract systems. In general however, these can be
categorised in two basic concepts. The first and by far the most common is the construction of a plenum
within the tunnel roof space, either from concrete, or by building a soffit from PROMATECT boards.
9.2 SUSPENDED CEILINGS, SMOKE EXTRACTION PLENUMS

A common way of providing smoke extraction systems in tunnels is the construction
of a smoke extraction plenum in the tunnel roof space. This is the transverse
ventilation system. In an emergency the smoke and hot gasses will be extracted into
the plenum through smoke inlets or hatches. Please refer to FIGURE 55 at right for two
typical examples.
FIGURE 55:
SMOKE EXTRACTION PLENUMS
Such a plenum can either be constructed out of concrete or steel. Regardless of the
selected construction method, the structural integrity of this plenum during fire is of
utmost importance as the ventilation philosophy is depending on it. In case the
plenum (or part of it) collapses during the event of fire, the intended smoke
management approach will be lost. All possible major implications are the usual result,
not to mention the hampering effect they would have on emergency response teams.
A smoke extraction plenum of this nature gets exposed to tunnel fire temperatures
from both sides, i.e. from below but also from the top because hot gases are pulled
into the duct. Temperature exposure will be equally high from both sides, especially at
the location of the hatches near the fire source.
Regardless of the selected construction method, such a plenum system therefore
requires thermal protection from both sides, not only from below.
Where a plenum is to be constructed out of concrete, please refer to SECTION 8 in this
manual outlining concrete protection.
The other option is to construct the plenum using a steel frame, spanning from wall to wall, with
intermediate hanger rods if mechanically required. As described above, such a frame requires thermal
protection from below and from above. The hanger rods also require thermal protection in order to
prevent elongation due to thermal expansion, which has the potential to cause unwanted deflection and
sagging of the plenum.
55
The steel frame should be designed so that it can cope with:

Dynamic load cycles coming from passing traffic.
Additional weight of the PROMATECT boards, taking into account the potential additional
weight of water which may be absorbed into the boards.
Elevated temperatures in case of fire and still retain its function.
Typically the design of such a suspended steel frame is conducted by a local structural engineer. Promat
has designed smoke extraction plenum systems for use in tunnels and have fire tested a number of
different configurations using PROMATECT-T boards.
SYSTEM COMPONENTS
The smoke extraction plenum system consists of a load bearing steel frame, which can be made out of
square hollow sections (SHS) or trapezoidal steel decking. The PROMATECT-T boards are screwed to
either side of the steel frame, also covering the edges at the framing exposed at the location of hatches.
The amount of steel used in terms of kilograms per square metre surface has an effect on the
temperature development within the plenum system. The more kilograms of steel per square metre, the
better the heat sink (heat absorption), hence the lower the temperatures will be on the steel. In contrast,
if less steel is required for structural reasons it should be noted that the temperature development will
increase as a function of time.
The fire tests took this effect into account, varying the heat sink effect, in combination with the selected
PROMATECT-T thickness. In this way Promat, has developed a design model in which these
parameters are fully considered. For example, when exposed from both sides by the RWS fire curve,
steel temperatures have been recorded between 285C and 570C, depending on the thickness of the
PROMATECT-T boards and the mass of steel being used.
On the basis of the test data generated by these fire tests reports, the structural engineer can design the
loadbearing steel frame, with Promat advising on the required thickness of the PROMATECT-T boards.
Temperature development rates on individual component members are available at Promat offices.
9.3 CABLE PROTECTION SYSTEMS

In the event of a fire it is vital to the safety of tunnel occupants that certain electrical systems remain
functioning until people have escaped. Such systems therefore require protection from fire for a
specified period of time and include:
Lighting for means of egress (emergency escape route lighting) and areas of refuge,
Exit signs,
Communications,
Electrically operated extinguishing systems,
Electrically operated fire and smoke alarms,
Ventilation and smoke extraction systems,
Tunnel drainage and fire pumps.
In addition to protection from fire outside the duct, it is normally vital that any fire within the duct is
contained, e.g. if cable sheathing ignites due to an electrical overload.
A suitably designed duct will:
1) Prevent the propagation of fire from one compartment to another;
2) Assist in maintaining escape routes;
3) Ensure the continuing operation of services;
4) Reduce damage to localised areas;
5) Contain smoke and toxic fumes from burning cables if the fire was within the cable enclosure.
By enclosing standard cables in the Promat cable duct systems, all the above requirements can be met,
providing up to 240 minutes fire protection, depending on the duct construction, and the fire exposure
curve. This avoids the use of more expensive and bulkier fire-rated cables, which cannot provide
performance to the more extreme exposure curves, such as the HCM and RWS fire curves.
56
DESIGN CONSIDERATIONS
The following points are some of the factors which should be considered when determining the correct
specification to ensure the cable duct system provides the required fire performance.
1) APPLICABLE TIME-TEMPERATURE CURVE
2) MAXIMUM ALLOWABLE TEMPERATURE ON THE SPECIFIC CABLE(S)
Non fire rated cables can generally operate in temperatures of approximately 130-150C for
short periods of time. However, such increases in cable temperatures do increase the electrical
resistance of the cable. The former temperatures are regularly used as the performance design
criteria for fire rated cable protection systems in tunnels. It should be noted that the majority of
fibre optic cables begin to break down once exposed to temperatures in the range of 50-80C.
FIGURE 56:
SERVICE TUNNEL
3) THE CROSS SECTION OF THE ENCLOSURE

The larger the perimeter of the enclosure around the cables, the greater the area exposed to fire
and thus more heat enters the duct. In instances where a three sided duct is constructed with
the fourth side being the concrete structure, the concrete will act as a heat sink, which will delay
the increase in the air temperature inside the duct. This in turn will ensure functionality of the
cables for a greater duration as the rise in cable temperature is postponed.
4) THE AMOUNT OF COPPER/ALUMINIUM WITHIN THE CABLE DUCT
The biggest heat conductor in a cable protection system is the copper/aluminium wire core itself.
Although the protection system provides thermal protection to the cables, the heat sink effect
into the cables themselves can be rather large as the cables are heated from ambient
temperatures to a maximum of approximately 130-150C. The greater the volume of
copper/aluminium wire within the enclosure, the greater the heat sink effect and the
longer functionality can be sustained. If a lower volume of cable is used, thicker fire protection
systems may be required.
5) REQUIRED FIRE PERFORMANCE
Generally, the most onerous requirement is to maintain the integrity of the circuit(s) when the
system is exposed to external fire. This means the cables must continue to function at full
capacity whilst exposed to fire. If this continued functionality is not required, the performance
specification may be reduced by the approval authority to provide only stability, integrity and
insulation of the duct system itself and/or the wall and floor penetrations.
6) SUPPORTING STRUCTURE
The supporting hangers and their fixings should be capable of bearing the load of the complete
cable system including any applied insulation material or other services suspended from it.
Chemical anchors are not generally suitable. It is usually not advisable to use unprotected
hangers if the stress exceeds 6N/mm2 and/or if hanger lengths exceed 2000mm. Unprotected
hangers are not allowed where they may be exposed to the RWS or HCM fire curve. It should
be noted that even stainless steel hangers will not survive such a thermal attack for long,
regardless of the stainless steel grade or tension stress level. The use of protected hangers is
therefore advisable.
7) PENETRATIONS THROUGH WALLS AND FLOORS
Care should be taken to ensure that movement of the cable system in ambient or in fire
conditions does not adversely affect the performance of the wall, soffit or floor or any penetration
seal. Also the sagging of the cable duct as a result of the elongation of the hangers must be
addressed at the penetration point of the duct through any wall, floor etc.
8) OTHER REQUIREMENTS
Acoustic performance, thermal insulation, water tolerance, strength and appearance can also be
important considerations.
Promat has conducted extensive fire testing on cable protection systems for tunnel applications, using
PROMATECT-H and PROMATECT-T boards. On the basis of the test data obtained, a design guide
has been developed taking the above mentioned parameters into account, thus balancing the
requirements to optimise the required PROMATECT thickness. These systems have been designed to
cope with the most severe time-temperature curve applied in tunnel design, the RWS fire curve. For
details of systems exposed to ISO Cellulosic, Standard Hydrocarbon or RABT time/temperature curves,
please consult the nearest local Promat office.
See FIGURE 57 at the right of this page for some examples of cable protection systems constructed from
PROMATECT boards.
FIGURE 57 (four examples) :

PROMAT CABLE PROTECTION SYSTEMS
57

9.4 SAFE HAVENS
In long tunnels, safe havens should form an integral part of the tunnel design. Recent fires in tunnels
have shown that exposure to smoke and toxic fumes from burning vehicles is the main cause of loss
of life. Deaths occur even at relatively short distances from the seat of the fire. The provision of safe
havens therefore is imperative in long tunnels, both to provide protection for passengers from vehicles
until fire and emergency personnel can reach them and also as a place which can provide respite from
heat and smoke for fire fighters.
Ideally, any safe haven should have a minimum fire resistance period to match that of the main structural
protection, and should be constructed in such a manner that is resistant to both heat (insulation) and
ingress of smoke into the chamber. In recent fires, some personnel who have managed to reach a safe
haven but have then succumbed through exposure to the effects of heat and smoke ingress into the
chamber. Consideration should therefore be given to providing a separate air supply for these areas.
FIGURE 58 (three examples) :
STEEL FIRE DOORS
WITHIN TUNNELS
Promat can offer the designs and systems required to construct such safe areas for all types and
durations of fire exposure. Please contact Promat for further details.
9.5 FIRE DOORS

Fire rated doors within tunnels are installed to provide a means of egress and to prevent the spread of
fire, hot gases and smoke from the tunnel to the surrounding compartments. Fire doors are installed:
at cross connections between two tunnel tubes,
to provide access to an escape route (mid tunnel channel in an immersed tunnel),
to protect people who have entered safe havens.
In view of the smoke emissions from vehicles, and the high toxicity of this smoke as a result of the
types of materials used in modern car manufacture, it is also imperative that doors provide a high
degree of resistance to the passage of smoke. Ideally, where used as access to safe havens,
doors should provide a high degree of thermal insulation to reduce the affects of heat on the occupants
of the chambers.
Any fire door situated within a tunnel should be capable of providing the same degree of corrosion
resistance to the aggressive and polluted environment of a tunnel as any other services.
In the design phase of a fire door, it should be noted that elongation of steel members will cause gaps
around the perimeter of the door, potentially introducing failure of the system. In addition to elongation,
steel members also tend to curve as a result of heating on one side only.
A tunnel fire door should be fire tested in two configurations:
1) The door leaf opening away from the heat source;
2) The door leaf opening into the heat source.
Promat has designed a fire rated door suitable for tunnels and fire tested according to the RWS
standard. For ease of operation in an emergency situation, the leaf is designed as a sliding door which
requires a minimum of force to open. The door provides thermal insulation for the full duration of
120 minutes when exposed to the RWS fire development. It also retains its integrity.
The material used to provide insulation to the door leaf is PROMATECT-T. The perimeter of the door
leaf, as well as some connections to the surrounding walls, are sealed by means of PROMASEAL-PL
intumescent strips.
58
10. FAQs
1) SHOULD VERTICAL TUNNEL WALLS BE PROTECTED AS WELL?
This depends on the assessment of the risk by the relevant authorities and fire consultants. In many tunnels up to
1m of the wall down from the tunnel soffit requires fire protection. Recent research (Runehamar) suggests that walls
do need some degree of protection.
2) WHICH IS THE BEST PROTECTION METHOD TO MINIMISE SERVICING REQUIREMENTS AFTER THE TUNNEL IS COMPLETED
POST CLADDING IS EASIER TO REMOVE, WHILE LOST SHUTTERING IS MORE DIFFICULT?
Promat has over 32 years of experience detailing and providing fire protection systems to tunnel applications. To
date there has not been a requirement to totally remove PROMATECT boards for servicing. It is true that post
cladding facilitates ease in retrieval.
3) HOW ARE THE CRACKS IN THE CONCRETE DURING FIXING OF PROTECTION MATERIAL TREATED?
Cracks in concrete pose no problem to the PROMATECT boards. If cracks in the concrete need to be repaired,
the boards can be removed, or drilled through to gain access to the concrete for grouting repairs.
4) DOES PROMATECT PROTECTION INHIBIT REGULAR INSPECTION AND MAINTENANCE PROCEDURES OF THE TUNNEL,
ESPECIALLY FOR WATER SEEPAGE AND CONCRETE SPALLING?
Water seepage is expected especially in sub-sea tunnels such as those in the Netherlands. For example,
Westerschelde Tunnel has a 12m water column. PROMATECT can be soaked by water seepage but the boards
are unaffected by water. Wet spots are therefore visible and hence do not inhibit inspection.
5) HOW ABOUT REBAR CARBONISATION? HOW WOULD A PROMATECT LINING AFFECT TREATMENT OF THIS PROBLEM IN TUNNELS?
The concrete cover should be designed for addressing this aspect, although the PROMATECT lining shields the
concrete from direct contact of aggressive car pollution. An examination of a nine year old PROMATECT board
cladding to Velser Tunnel in the Netherlands showed negligible loss of strength. No rebar carbonisation was visible
in the concrete.
6) WHAT IS THE EXPERIENCE OF SUCH REPAIRS IN OTHER PROTECTED TUNNELS?
The worst case scenario is the PROMATECT panel has to be removed to allow access for concrete repair. This is
quite easily achieved.
7) HOW WILL PROTECTION MATERIAL REACT TO CHEMICALS IN WATER SEEPAGE?
PROMATECT boards are inert and will not have adverse reaction to chemicals in the water.
8) HOW WILL PROTECTION MATERIAL REACT TO ALTERNATING PRESSURE FROM VEHICULAR TRAFFIC?
A test has been carried out in Germanys iBMB subjecting PROMATECT specimens of alternating pressures, three
times more cycles than normally encountered in vehicular tunnels. No displacement of the system occurred.
9) HOW DO WE BUILD IN MAINTENANCE AND SERVICE PROCEDURE FOR PROMATECT AFTER PROTECTED TUNNEL IS OPERATIONAL?
PROMATECT boards require little or no maintenance, other than a visual inspection.
10) HOW DOES FIXING OF SERVICES AND LIGHTING TO A PROTECTED CONCRETE SOFFIT AFFECT FIRE PERFORMANCE
OF THE CONCRETE?
Drilling through the panels does not adversely affect the performance of the system, assuming of course that the
installer does not go too far and drills holes everywhere. Tests have been carried out to both RWS and Hydrocarbon
curves where services have been bolted through the PROMATECT (simulated in the tests by suspending weights
from expansion bolts) and the performance of the system is consistent between these tests and the standard tests
where no penetrations have been made. Of course, all services should be supported directly from the concrete and
the installer should not rely on fixing any services only to the PROMATECT boards.
11) HOW DO WE ENSURE THE SCREWS OR BOLTS REMAIN IN SITU?
If PROMATECT is used as permanent shuttering, screws are embedded within the concrete and thus cannot fall
out. If the bolts used for fixing using the post installation method described in SECTION 8, are not tight, the board
will fall as the support is removed.Tests have been carried out to show that even without the screws, a section of
board used as shuttering has very high adhesion to the concrete and will not fall away. Tests on fully soaked boards
have been carried out to simulate the effects on suction and to ascertain whether the bolt heads and washers
would pull through the board. Tests were carried out on 15mm, 20mm, 25mm and 30mm showed that very high
loads are required to pull the fixings through the boards. The average pull through strength measured for a 25mm
board, fully immersed in water for 72 hours prior to test, was a pull through load of 1884N for a 6mm diameter
expansion bolt and 1271N for a 5mm diameter screw.
12) WHAT HAPPENS IF THERE ARE ANY POST INSTALLATION GAPS BETWEEN THE PROMATECT BOARDS?
This depends on the size of the gaps. Panels have been tested where gaps of 3mm were deliberately left between
the panels in an attempt to simulate poor installation. No adverse affects were recorded in these tests.
59
11. TUNNEL FIRE RESEARCH

In recent years a lot of research has been conducted, mainly under the auspices of the European
Union.The results of this research will eventually translate into directives, guidelines and standards for
tunnel fire safety around the world.
11.1 FIT
FIT is the abbreviation for European Thematic Network on Fire in Tunnels. FIT provides a European
platform for dissemination of information of up-to-date knowledge and research on Fire & Tunnels. FIT
represents 33 members from 12 European Countries.
To optimise benefits of the knowledge throughout Europe from real fire accidents, testing and research
there are many benefits to using all available information via a European Thematic Network. The
following main objectives have been identified for the FIT Thematic Network:
1) The network dissemination of RTD and design results obtained in European and National RTD
projects. The aim is to optimise research efforts, to reach critical mass and to enhance impact
at a European level by combining the results of the different projects.
2) FIT will establish a set of consultable databases with essential knowledge on fire in tunnels.
3) Realise recommendations on design fires for tunnels.
4) To develop European consensus for fire safe design on the basis of existing national regulation,
guidelines, code of practices and safety requirements.
5) Define best practices for tunnel authorities and fire emergency services on prevention and
training, accident management and fire emergency operations.
11.2 DARTS
DARTS is an RTD-project on Durable and Reliable Tunnel Structures.
The project was conducted during 2001-2004 by a partnership of eight European companies. The
DARTS-project is performed with financial support of European Communities under the Fifth
Framework Programme, Competitive and Sustainable Growth Programme (GROWTH 2000).
The objective of the DARTS-project is to develop operational methods and supporting practical tools
for the best proactive decision-making process for selecting in each individual case, the cost optimal
tunnel type and construction procedures regarding environmental conditions, technical qualities, safety
precautions and long service life.
DARTS is developed for the main current types of tunnels: rock tunnels, bored tunnels, NATM tunnels,
immersed tunnels and cut and cover tunnels.
11.3 UPTUN
UPTUN is the acronym for Cost effective, Sustainable and Innovative Upgrading Methods for Fire Safety
in Existing Tunnels, a European RTD-project funded by the European Commission in FP5.
The main UPTUN project objectives are:
1) To develop innovative technologies where appropriate and relevant, comparing to and
assessing existing technologies for tunnel application. Focus is on technologies in areas of
detection and monitoring, mitigating measures, influencing human response, and protection
against structural damage.
2) To develop, demonstrate and promote procedures for rational safety level evaluation, including
decision support models and knowledge transfer.
60
In order to achieve these objectives overall, a strong European consortium was needed, covering all
relevant expertise, with sufficient mass and impact to ensure adoption of UPTUN deliverables
throughout Europe. The consortium was built around prominent tunnel safety institutes in Europe,
balancing owners, industry, research and other stakeholders on the one hand with the (tunnel) member
states on the other.
The UPTUN consortium consists of 41 members from 13 different EU member states, one EEA member
state and three accession countries. The distribution of the input to the project was well balanced over
the eastern, northern, southern and western EU member states.
The project was specifically targeted at ensuring a pan European approach towards improvement of
fire safety in European tunnels. This will enable European tunnel operators and regulators to benefit
from economies of scale resulting from a European approach and also create additional added-value
for the community.
Among others, a full scale fire test in the Runehamar tunnel in Norway was conducted in the framework
of UPTUN.
11.4 SIRTAKI
SIRTAKI Safety Improvement in Road & rail Tunnels using Advanced ICT and Knowledge Intensive
DSS is a IST Project supported by the Commission of the European Communities in the framework
of the Key Action I of IST Programme.
The strategic goal of SIRTAKI is the development and assessment of an advanced tunnel
management system that specifically tackles safety issues and emergencies and integration within
overall network management.
A multidisciplinary consortium with representation from all participating members, including local
authorities, system providers and research institutions from eight different European countries, has
implemented numerous SIRTAKI initiatives over 36 months from September 2001.
61
APPENDIX 1: PHYSICAL PROPERTIES

TABLE 16: PROMATECT-H PHYSICAL PROPERTIES
Properties
Description
Neutral designation
Calcium silicate matrix, asbestos free
Material class
Non combustible in accordance to DIN4102,

EN 13501-1 (A1) and BS476: Part 4.
Surface spread of flame
Class 1 in accordance to BS476: Part 7
Building regulations classification
Class 0
Bulk density (air dry)
900kg/m3 (nominal)
Thermal conductivity ()
0.17W/mK at 20C (typical value)
Alkalinity
Nominal pH 12
Water vapour diffusion

resistance factor ()
20 (tabulated value)
Moisture content
Air-dried, approximately 7%
Water absorption capacity
Maximum 0.55g/cm3
Dimensions and tolerances

(for squared boards)
For sheets of 1220mm x 1220mm:

Length x width 0.5mm
Squareness
1mm across diagonals
Thickness
27mm 0.5mm

(for standard boards)
For sheets of 1250mm x 2500mm

and 1250mm x 3000mm:
Length x width 3mm
Thickness
27mm 1.5mm
Surface condition of
standard boards
Visible face smooth, opposite face honeycombed.
Biological
Inorganic material that will not rot and not attrack pests.
Flexural strength, F
Longitudinal: 10N/mm2 (average production value)

Transverse: 5.5N/mm2 (average production value)
Tensile strength, T
Longitudinal: 5N/mm2 (typical value)

Transverse: 4N/mm2 (typical value)
Compressive strength
(perpendicular to surface of board)
9.3N/mm2 (typical value)
Screw pull out resistance
Screw inserts (Type B 3815) RAMPA

Screw depth of 15mm on board face: 330N (typical value)
All physical and mechanical property values are averages based on standard production and tested
according to internal procedures. The typical values are given for guidance. The figures can change
dependent on the test methods used. If a particular value is of prime importance or a specification,
please consult Promat Technical Department.
62
TABLE 17: PROMATECT-T PHYSICAL PROPERTIES

Properties
Description
Neutral designation
Matrix engineered calcium silicate-aluminate
Material class
Non combustible in accordance to DIN4102,

EN 13501-1 (A1), BS476: Part 4.
Surface spread of flame
Class 1 in accordance to BS476: Part 7
Building regulations classification
Class 0
Bulk density (ovendry)
900kg/m3 (nominal)
Thermal conductivity ()
0.21W/mK at 20C (typical value)
Alkalinity
Nominal pH 10
Water vapour diffusion

resistance factor ()
5 (typical value)
Water absorption capacity
Maximum 0.6g/cm3

(for squared boards)

Length x width 0.5mm
Squareness
1mm across diagonals
Thickness
15mm, 20mm, 25mm, 30mm,
35mm and 40mm 0.5 mm

(for standard boards)

Length x width 3mm
Thickness
15mm, 20mm, 25mm, 30mm,
35mm and 40mm 0.5 mm
Surface condition of
standard boards
Visible face smooth, opposite face honeycombed.
Biological
Inorganic material that will not rot and not attrack pests.
Flexural strength, F
Longitudinal: 4.5N/mm2 (average production value)
Tensile strength, T
Longitudinal: 1.2N/mm2 (typical value)
Compressive strength
(perpendicular to surface of board)
1% deformation: 1.2N/mm2 (typical value)

10% deformation: 7.8N/mm2 (typical value)
Screw pull out resistance
20mm deep air dry:

657N (typical value)
20mm deep saturated: 372N (typical value)
(quick fix screw 5mm x 50mm)
Bolt pull through resistance
For 25mm boards: 3.22N (typical value)

(bolt M8, washer 30mm)
All physical and mechanical property values are averages based on standard production and tested
according to internal procedures. The typical values are given for guidance. The figures can change
dependent on the test methods used. If a particular value is of prime importance or a specification,
please consult Promat Technical Department.
63
APPENDIX 2: WORLDWIDE TUNNEL PROJECT REFERENCE

TABLE 18: WORLDWIDE TUNNEL PROJECT REFERENCE
Year
Country
City/Location
Project
Type of tunnel
2008-2009
2007-2008
Germany
Limburg
Schiedetunnel
Road tunnel
Germany
Frankfurt am Main
Theatertunnel
Road tunnel
2007
Australia
Brisbane
INB1 Tunnel
Busway
2007
China
Suzhou
Suzhou Dushu Lake Tunnel
Road tunnel
2007
France
Monaco
Monaco Tunnel
Road tunnel
2007
Germany
Dsseldorf
Werstener Tunnel
Road tunnel
2007
Germany
Hamburg
S-Bahn-Tunnel, Flughafen
Suburban railway
2007
Germany
Berlin
Bundesplatztunnel
Road tunnel
2007
Germany
Ettlingen bei Karlsruhe
Wattkopf-Tunnel
Road tunnel
2007
UK
Glasgow
Clyde Tunnel
Road tunnel
2007
United Arab Emirates
Abu Dhabi
Abu Dhabi Airport Tunnel
Road tunnel
2007
United Arab Emirates
Dubai
Palm Jumeirah Tunnel
Road tunnel
2006-2007
Australia
Sydney
Epping-Chatswood Rail Link
Rail tunnel
2006
France
Paris
Porte des Lilas
Road tunnel
2006
Germany
Hamburg
U-Bahn-Tunnel (U 4) Gnsemarkt
Underground railway
2006
Singapore
From East Coast Parkway

to Tampines Expressway
Kallang Paya Lebar Expressway
Road tunnel
2006
Spain
Madrid
PIO XII
Road tunnel
2005-2006
Australia
Sydney
Lane Cove Tunnel
Road tunnel
2005
Austria
Vienna
Absberg Tunnel
Road tunnel
2005
China
Nanjing
Nanjing Jiuhua Mountain Tunnel
Road tunnel
2005
France
73 Novalaise
Tunnel de IEpine
Road tunnel
2005
France/Italy
Frejus
Frejus Tunnel
Road tunnel
2005
Italy
Tindari
Galleria Tindari
Rail tunnel
2005
Italy
Lecco
Lecco Tunnel
Road tunnel
2005
Germany
Stuttgart
Messetunnel
Road tunnel
2005
Netherlands
Abcoude
Aquaduct Abcoude
Rail tunnel
2005
Netherlands
Roermond
Roer Tunnel
Road tunnel
2004-2005
Australia
Sydney
Cross City Link Tunnel
Road tunnel
2004
China
Shanghai
Shangphai Outer Ring Tunnel
Road tunnel
2004
France
09 Foix
Tunnel de Foix
Road tunnel
2004
France
34 Lodve
Tunnel de la Vierge
Road tunnel
2004
France
66 Porta Hospitalet
Tunnel de Puymorens
Road tunnel
2004
France
73 Moutier
Tunnel de Siaix
Road tunnel
2004
Germany
Hornberg (Schwarzwald)
Hornberg-Tunnel (B 33)
Road tunnel
2003-2007
France
73 Chambry
Tunnel des Monts
Road tunnel
2003-2005
France/Italy
Frejus
Frejus Tunnel
Road tunnel
2003-2004
France
A40 Nantua
Tunnel de Chamoise
Road tunnel
2003
Australia
Sydney
Central Business District
Road tunnel
2003
Australia
Brisbane
INB3 Tunnel
Service tunnel
2003
China
Huangzhou
XiHu Lake Tunnel
Road tunnel
2003
France
Toulouse
Tunnel Metro de Toulouse
Road tunnel
2003
Netherlands
Terneuzen
Westerschelde Tunnel
Road tunnel
2003
Sweden
Gothenburg
Gotha Tunnel
Road tunnel
2002
Germany
Freiburg
Schtzenallee-Tunnel (B 31)
Road tunnel
64
65
Year
Country
City/Location
Project
Type of tunnel
2002
Australia
2002
Austria
Sydney
M5 Tunnel
Road tunnel
Vienna
Rennweg train station
Train station
2002
2002
Austria
Vienna
Sanki Marx train station
Train station
China
Nanjing, Jiangsu
Xuan Wu Lake Tunnel
Road tunnel
2002
Denmark
Copenhagen
Copenhagen Metro Station
Metro tunnel
2002
France
Toulon
Toulon Tunnel
Road tunnel
2002
France
46 Valroufie
Tunnel de Constans
Road tunnel
2002
France
46 Pinsac
Tunnel de Terregay
Road tunnel
2002
Japan
Tokyo
Tokyo Port seaside tunnel
Road tunnel
2002
Netherlands
Amsterdam
A5 Schiphol Airport, airplane viaduct
Road tunnel
2002
Netherlands
Rotterdam
Caland Tunnel
Road tunnel
2002
Netherlands
Roelofarendsveen
High Speed Line Aquaduct
Rail tunnel
2002
Netherlands
Rotterdam
High Speed Line Oude Maas & Dordtse Kil
Rail tunnel
2002
Netherlands
Dordrecht
Kil Tunnel
Road tunnel
2002
Netherlands
Voorburg
Seitwende
Road tunnel
2001
Australia
Sydney
Eastern Distributor
Road tunnel
2001
China
Ningbo, Zhejiang
Ningbo river crossing tunnel
Road tunnel
2001
France
74 Chamonix
Tunnel du Mont Blanc
Road tunnel
2001
Germany
Hamburg
Elbtunnel, Western Tube
Road tunnel
2001
Germany
Hamburg
Elbtunnel, 4th Tube
Road tunnel
2001
Japan
Tokyo
Rinkaidoro
Road tunnel
2001
Netherlands
Rotterdam
1e Benelux Tunnel
Road under canal
2001
Netherlands
Rotterdam
2e Benelux Tunnel
Road tunnel
2001
Netherlands
Enkhuizen
Naviduct Enkhuizen
Road under lock
2001
Netherlands
Oud Alblas
Sophiatunnel
Underground route
2000-2001
Germany
Hamburg
Elbtunnel, Western Tube
Road tunnel
2000
Australia
Melbourne
Burnley Tunnel
Road tunnel
2000
France
73 Saint Michel de
Maurienne
Tunnel dOrelle
Road tunnel
2000
Germany
Hamburg
Elbtunnel, Central Tube
Road tunnel
2000
Germany
Hamburg
Elbtunnel, Eastern Tube
Road tunnel
2000
Netherlands
Leidschendam
Aquaduct onder de Vliet
Tramway route
1999
Germany
Hamburg
Krohnstiegtunnel
Road tunnel
1999
Germany
Freiburg
Schtzenalleetunnel
Road tunnel
1999
Japan
Tokyo
Dainikouro
Road tunnel
1999
Netherlands
Rotterdam
2e Beneluxtunnel
Road under canal
1999
Netherlands
Rotterdam
Botlektunnel
Road under canal
1999
Netherlands
Amsterdam
Ij-tunnel
Road under canal
1999
Netherlands
Leidschendam
Seitwendetunnel
Road tunnel
1999
Netherlands
Zeeland
Westerscheldetunnel
Road under canal
1998
Australia
Perth
City Northern Bypass Tunnel
Road tunnel
1998
Switzerland
Grellingen
Eggflue-Tunnel
Road tunnel
1997
China
Hong Kong
Hong Kong International Airport Tunnel
Road tunnel
1997
Germany
Bad Godesberg
Urban tunnel, below the B9
Road tunnel
TABLE 18: WORLDWIDE TUNNEL PROJECT REFERENCE Continued from page 65

Year
Country
City/Location
Project
Type of tunnel
1997
Netherlands
Alphen/Rhine
Aquaduct Alphen
Road tunnel
1997
Netherlands
Delft
Aquaduct Delft
Tramway route
1997
Netherlands
Amsterdam
Schiphol (kaagbaan)
Road under runway
1997
Netherlands
Schiphol
Schipholtunnel
Road tunnel
1996
Netherlands
Akrum
Aquaduct Akrum
Road tunnel
1995
Singapore
Marina Centre
Suntec City Convention Centre
Underground
parking facility
1994
Belgium
Brussels
Belliard Tunnel
Road tunnel
1994
Germany
Hamburg
Elbtunnel
Road tunnel
1994
UK
London
Leicester Square
Electricity substation
1994
Italy
Mont Blanc
Mont-Blanc-Tunnel
Road tunnel
1994
Malaysia
Shah Alam
Shah Alam Sports Complex
Road tunnel
1994
Netherlands
Barendrecht
Heineoordtunnel
Road under canal
1994
Netherlands
Velsen
Wijkertunnel
Road under canal
1993
China
(formerly UK)
Hong Kong
Times Square Shopping Complex
Underground
parking facility
1993
UK
Wadham
Power transmission tunnel
Service tunnel
1993
Malaysia
Kuala Lumpur
Denmark House
Underground
parking facility
1993
Malaysia
Shah Alam
Shah Alam Sports Complex
Underground
parking facility
1993
Malaysia
Kuala Lumpur
Sogo Department Store
Underground
parking facility
1993
Netherlands
Schiphol
Schipholtunnel 2
Road under runway
1993
Netherlands
Ijmuiden
Wijkertunnel
Road under canal
1993
Singapore
Orchard
Ngee Ann City
Underground
parking facility
1992-1998
China
Hong Kong
Hong Kong MTR
Underground stations
1992
China
(formerly UK)
Hong Kong
Route 5
Road tunnel
1992
Germany
Munich
Munich Airport, tunnel
Suburban railway
1992
UK
London
Bow Road Station
Underground station
1992
UK
London
Eurostar Waterloo
International rail terminal
1992
Malaysia
Kuala Lumpur
Swiss Garden Hotel
Underground
parking facility
1992
Netherlands
Grouw
Aquaduct Grouw
Road tunnel
1992
Netherlands
Schiphol
Schipholtunnel
Road under railway
1991
Belgium
Antwerp
Bevrijdingstunnel
Road tunnel
1991
Netherlands
Zeeland
Vlaketunnel
Road under canal
1990
Belgium
Antwerp
Beveren Tunnel
Road tunnel
1990
Belgium
Zelzate
Hoge Weg Tunnel
Road tunnel
1990
Belgium
Antwerp
Liefkenshoek Tunnel
Road tunnel
1990
Belgium
Antwerp
Tijsmans Tunnel
Road tunnel
1990
China
(formerly UK)
Hong Kong
Pacific Place
Road under canal
66
Year
Country
City/Location
Project
Type of tunnel
1990
UK
London
St. Pauls Thames Link
Underground station
1990
Netherlands
Barendrecht
Heineoordtunnel
Road under canal
1990
Netherlands
Velsen
Velsertunnel
Road under canal
1989-1992
Singapore
Singapore
MRTC
Underground stations
1989
Australia
Sydney
Sydney Harbour Tunnel
Road tunnel
1989
Belgium
Brussels
Leopold II Tunnel
Road tunnel
1989
China
(formerly UK)
Hong Kong
Second Cross Harbour Tunnel
Road tunnel
1989
China
(formerly UK)
Hong Kong
Eastern Harbour Crossing
Road tunnel and

suburban railway
1989
Germany
Bad Ems
Ems Tunnel
Road tunnel
1989
Netherlands
Schiphol
Schipholtunnel
Road tunnel
1989
Netherlands
Schiphol
Schipholtunnel
Road under runway
1989
Netherlands
Hendrik Ido Ambacht
Tunnel onder de Noord
Road tunnel
1989
Netherlands
Amsterdam
Zeeburgertunnel
Road tunnel
1989
USA
Boston
Harbour Tunnel CANA
Road tunnel
1988
Belgium
Antwerp
Kennedy Tunnel
Road tunnel
1988
UK
Medway
Power transmission tunnel
Service tunnel
1987
China
(formerly UK)
Hong Kong
First Cross Harbour Tunnel
Road tunnel
1987
Netherlands
Schiphol
Schipholtunnel
Road under runway
1987
Switzerland
Genf
Suburban railway
Suburban railway
1986
Belgium
Antwerp
Jan de Voslei Tunnel
Road tunnel
1986
Netherlands
Schiphol
Schipholtunnel
Road under runway
1985
Belgium
Brussels
Rogier Tunnel
Road tunnel
1982
Belgium
Brugge
Tunnel t Zand
Road tunnel
1981
Belgium
Antwerp
Craeybeckx
Road tunnel
1980
Germany
Berlin
Schlangenberger
Road tunnel
1980
Switzerland
St. Gotthard
Gotthard Tunnel
Road tunnel
1975
Germany
Hamburg
Elbtunnel
Road tunnel
1963
UK
Dartford
Dartford Tunnel
Road tunnel
The information contained within this table is believed to be accurate at the time of preparation of this document. Latest and
complete information on this worlwide tunnel project reference can be obtained from www.promat-tunnel.com.
67
Promat Worldwide
An Extensive Global Service Network
HEAD OFFICE, BELGIUM
DENMARK
POLAND
Promat International N.V.

Bormstraat 24
2830 Tisselt
Tel:
+32-15 718 100
Fax: +32-15 718 109
Email: [email protected]
www.promat-international.com
www.promat-tunnel.com
Promat ApS
Tstrup Stationscenter
Selsmosevej 2, 1. floor, 2630 Tstrup
Tel:
+45 7020 0482
Fax: +45 7020 0483
www.promat-nordic.com
Promat TOP Sp. z o.o

ul. Przeclawska 8
03 879 Warszawa
Tel:
+48-22 212 2280
Fax: +48-22 212 2290
www.promattop.pl
FRANCE
RUSSIA
AUSTRALIA
Promat S.A.S.
BP 66 Rue de LAmmandier
78540 Vernouillet
Tel:
+33-1 3979 6160
Fax: +33-1 3971 1660
www.promat.fr
Promat GmbH
Reprsentanz
ul. Kantemirovskaya 58
115477 Moscow
Tel:
+7-495 325 4268
Fax: +7-495 231 7977
www.promat.ru
Promat Australia Pty. Ltd.

1 Scotland Road, Mile End South
SA 5031 Adelaide
Tel:
+61 1800 PROMAT
Fax: +61-8 8352 1014
www.promat-ap.com
Promat GmbH
Scheifenkamp 16
40878 Ratingen
Tel:
+49-2102 493 0
Fax: +49-2102 493 111
www.promat.de

3/273 Williamstown Road
VIC 3207 Port Melbourne
Tel:
+61 1800 PROMAT
Fax: +61-3 9645 3844
www.promat-ap.com
HONG KONG
AUSTRIA
Promat Ges.m.b.H.
Seybelgasse 13
1230 Wien
Tel:
+43-1 865 4501
Fax: +43-1 865 3600
www.promat.at
BELGIUM
Promat N.V.
Kuiermansstraat 1
1880 Kapelle-op-den-Bos
Tel:
+32-15 713 351
Fax: +32-15 718 229
www.promat.be
CHINA
Promat China Ltd.
Room 503, Block B, Qi Lin Plaza
13-35 Pan Fu Road
510180 Guangzhou
Tel:
+86-20 8136 1167
Fax: +86-20 8136 1372
www.promat.com.cn
Promat North China
(Division of Promat China Ltd.)
Room 1507 Building 5, SOHO Xiandaicheng
No.88 Jianguo Road, Chaoyang District
100022 Beijing
Tel:
+86-10 8589 1254
Fax: +86-10 8589 2904
www.promat.com.cn
Promat International N.V. 07/2008
GERMANY

Unit 1, 175 Briens Road
NSW 2152 Northmead
Tel:
+61 1800 PROMAT
Fax: +61-2 9630 0258
www.promat-ap.com
Promat International (Asia Pacific) Ltd.

Room 1010, C.C. Wu Building
302-308 Hennessy Road, Wanchai
Tel:
+852 2836 3692
Fax: +852 2834 4313
www.promat.com.hk
INDIA
(India Representative Office)
610-611, Ansal Imperial Tower
C-Block, Community Centre
Naraina Vihar, Naraina, 110028 New Delhi
Tel:
+91-11 2577 8413
Fax: +91-11 2577 8414
www.promat-ap.com
SINGAPORE
Promat Building System Pte. Ltd.
10 Science Park Road, #03-14 The Alpha
Singapore Science Park II
117684 Singapore
Tel:
+65 6776 7635
Fax: +65 6776 7624
www.promat-ap.com
SPAIN
Promat Ibrica S.A.
C/ Velazquez, 41 - 2 Planta
28001 Madrid
Tel:
+34-91 781 1550
Fax: +34-91 575 1597
www.promat.es
SWITZERLAND
Promat AG
Stationsstrasse 1
8545 Rickenbach-Sulz
Tel:
+41-52 320 9400
Fax: +41-52 320 9402
www.promat.ch
ITALY
UNITED ARAB EMIRATES
Promat S.p.A.
Corso Paganini 39/3
16125 Genova
Tel:
+39-010 248 8411
Fax: +39-010 213 768
www.promat.it
Promat Middle East

Suite 1805, Floor 18
Festival City Tower
P O. Box 123945, Dubai
Tel:
+971-4-232 9780
Fax: +971-4-232 9781
www.promat.co.uk
MALAYSIA
Unit 19-02-01, Level 2 PNB Damansara
No.19 Lorong Dungun, Damansara Heights
50490 Kuala Lumpur
Tel:
+60-3 2095 5111
Fax: +60-3 2095 6111
www.promat-ap.com
Promat (Malaysia) Sdn. Bhd.
Unit 19-02-01, Level 2 PNB Damansara
No.19 Lorong Dungun, Damansara Heights
50490 Kuala Lumpur
Tel:
+60-3 2095 8555
Fax: +60-3 2095 2111
www.promat-ap.com
CZECH REPUBLIC
NETHERLANDS
Promat s.r.o.
Ckalova 22/784
16000 Praha 6 - Bubenec
Tel:
+420-2 2439 0811
Fax: +420-2 3333 3576
www.promatpraha.cz
Promat B.V.
Zonnebaan 36
3542 EE Utrecht
Tel:
+31-30 241 0770
Fax: +31-30 241 0771
www.promat.nl
1. Promats policy is one of continuous development. We therefore reserve the right to change designs and specifications without prior notice.
2. All products are subject to our Terms and Conditions of sale, copies of which are available upon request.
3. This publication and information, illustrations etc. contained herein are provided in good faith and does not accept responsibility for any ommissions or
misinterpretations of any of the information. This publication has no statutory authority, and therefore approved documents or the appropriate Building
Regulations must be consulted in all instances.
UNITED KINGDOM
Promat UK Ltd.
The Sterling Centre
Eastern Road, Bracknell
RG12 2TD Berkshire
Tel:
+44-1344 381 300
Fax: +44-1344 381 301
www.promat.co.uk
VIETNAM
(Vietnam Representative Office)
Room 606 Giay Viet Plaza
180-182 Ly Chinh Thang Street
Ward 9, District 3
Hochiminh City
Tel:
+84-8 931 5964
Fax: +84-8 931 5964
www.promat-ap.com
Your local Promat Supplier

Tunnel Brochure PDF

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

Tunnel Brochure PDF

Uploaded by

Copyright:

Available Formats

July 2008

Tunnel Fire Protection

2. WHY PROTECT TUNNELS?

3. TYPES OF FIRE EXPOSURE _______________________________________________________ 24-31

4.1 HEATING RATE AND INTERNAL STRESSES ________________________________________________ 32

5. CHOOSING FIRE PROTECTION MATERIALS

5.1 CONSIDERATIONS WHEN APPLYING PROTECTIVE MATERIALS __________________________________ 37

6.1 BORED TUNNELS USING CONCRETE SEGMENTS____________________________________________ 38

7. FIXINGS ________________________________________________________________________ 39-43

8. METHODS OF APPLYING BOARD PROTECTION _____________________________________ 44-53

9. FIRE PROTECTION OF ESSENTIAL SERVICES

9.1 AIR SUPPLY AND SMOKE EXTRACTION SYSTEMS ________________________________________ 54-55

10. FAQs _____________________________________________________________________________ 59

APPENDICES ______________________________________________________________________ 62-67

TABLES, GRAPHS & DIAGRAMS

FIGURE 1: EXAMPLES OF FIRE DAMAGE ________________________________________________________ 7

FIGURE 20: TENSION CRACK CORROSION IN 1.4401 ______________________________________________ 41

2. WHY PROTECT TUNNELS?

Hokuriku Tunnel, Japan

Second, there is the performance of the

Example 1 Moorfleet Tunnel, Germany

Example 2 Mont Blanc Tunnel, Italy

Example 3 Velsen Tunnel, Netherlands

Example 4 Channel Tunnel, UK

In terms of fire protection within tunnel and

2. WHY PROTECT TUNNELS?

RECOMMENDATIONS OF THE GROUP OF EXPERTS ON SAFETY OF ROAD TUNNELS

DIRECTIVE 2004/54/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

2. WHY PROTECT TUNNELS?

Continued from page 9

2.2 IMPLICATIONS OF THE EU DIRECTIVE TO TUNNEL STRUCTURES

2.2.1 RISK MANAGEMENT

THE RISK = FIRE

2.2.2 RISK ANALYSIS

RISK = FREQUENCY x CONSEQUENCE

HEAT RELEASE RATE (MW)

2. WHY PROTECT TUNNELS?

Continued from page 11

Mechanical model, truss elements

2D, mesh for thermal analysis

WP4 TUNNEL SYSTEM STRUCTURAL RESPONSE

2. WHY PROTECT TUNNELS?

Continued from page 13

2.2.4 ECONOMIC AND ENVIRONMENTAL IMPACT

THE BOARD SOLUTION

THE SPRAY APPLIED SOLUTION

TABLE 2: TUNNEL FIRE HISTORY

Soviet Military column,

Hydraulic oil leak

Line closed for

Lorry with trailer

Suicidal car driver

Collision with end

Tyre caught fire

Lorry with trailer

HGV loaded with

Lorry with drums of

HGV carrying 16 tons

Goods train and

TABLE 2: TUNNEL FIRE HISTORY Continued from page 17

Lorry with 12 tons flour

Oil leakage motor,

Kurt Schumacher Platz

T = 20 + 1080* (1 - 0.325e-0.167t - 0.675e-2.5t)