ASSOCIATION ITA INTERNATIONALE DES TRAVAUX EN SOUTERRAIN AITES INTERNATIONAL TUNNELING ASSOCIATION Towards an In Consultative Status, Category II with the improved use United Nations Economic and Social Council of underground http : //www.ita-aites.org space International Tunneling Association Association Internationale De Travaux En Souterrain GUIDELINES FOR STRUCTURAL FIRE RESISTANCE FOR ROAD TUNNELS DIRECTIVES POUR LA RESISTANCE AU FEU DES STRUCTURES DE TUNNELS ROUTIERS BY Working Group No.6 Maintenance and Repair o Groupe de Travail N 6 Entretien et Repaire May, 2004 GUIDELINES FOR STRUCTURAL FIRE RESISTANC FOR ROAD TUNNELS TABLE OF CONTENTS CHAPTER DESCRIPTION PAGE NO. 1. INTRODUCTION 1-1 2. DESIGN CRITERIA 2-1 3. LINING MATERIAL BEHAVIOR 3-1 4. TUNNEL CATEGORIES 4-1 5. STRUTURAL ELEMENTS 5-1 6. SUMMARY & RECOMMENDATIONS 6-1 7. TUNNEL FIRE HISTORY 7-1 8. BIBLIOGRAPHY 8-1 9. APPENDIX A 9-1 03/05/2005 1-1 1. Introduction 1.0 General Fire resistance of tunnel structures is an important issue. If it is not properly addressed, fire in a tunnel can result in loss of life to both tunnel users and the fire and rescue services. The result- ing economic losses for both the tunnel owner/operator and to the local economy and environ- ment can be catastrophic. This document is the result of a co-operative effort between the World Road Association (PIARC) Technical Committee on Road tunnel Operation (C 3.3) and its Working Group 6 Fire and Smoke Control, and the International Tunnelling Association (ITA) Working Group 6 Re- pair and Maintenance of Underground Structures. The purpose of this co-operative effort is to develop guidelines for resistance to fire for road tunnel structures. However, many of the issues addressed in these guidelines are relevant to rail tunnels and these will be referenced as part of this document at a later date. (Part B). An agreement for a co-operative effort has been established between both parties to develop a common approach to fire resistance for structures. PIARC and ITA deem that such a guideline is necessary for the protection of the public. In this co-operation, the role of PIARC has been to define the objectives of resistance to fire of road tunnel structures, i.e. to develop appropriate design fires (mainly as time - temperature curves) and specify the required resistance times. Re- cent developments emerging from full scale fire tests carried out as part of the EC Funded Re- search Project “UPTUN” have shown that fires in tunnels can be much more severe than previ- ously assumed (even with non-hazardous goods). For this reason, this document makes refer- ence to the time - temperature curves developed in these tests. It also references the relative heat release curves (RHR) developed from these full-scale fire tests as separate graphs. The role of ITA is to develop guidelines for techniques and materials to answer these structural require- ments and make tunnels and their ancillary structures more resistant to fire damage. 1.1 Scope The scope of this document is to provide recommendations for the design of underground struc- tures which function as tunnels and ancillary structures for the safe operation of road tunnels. These guidelines take into consideration the time - temperature curves as recommended by PIARC and develop suitable means and methods for the protection of the structures from col- lapse both during the fire event and during rescue operations. These guidelines also reference the actual time - temperature and Heat Release Rate curves generated from the full-scale fire tests carried out in the Runehamar tunnel in Norway in September 2003. This document is intended to be a guideline and is to be used for road tunnels only and not for rail, mass transit, or pedestrian tunnels. However the basic principles for the protection of tun- nels and underground structures may be applied to other types of structures; in such cases spe- cial consideration must be given to the particular application and its own unique operational and other site-specific elements. This guideline is for informational purposes only and applicable codes, standards and local regulations must be consulted for compliance to specific structural and life safety requirements of the locale in which the structure is located. 1.2 Reason for Developing Guidelines 03/05/2005 1-2 Over the last 10 years there have been a number of serious underground fires in road tunnels. These fires have caused extensive loss of life and severe collateral loss to the infrastructure. Aside from the tragic loss of life, the long-term financial effects to the local infrastructure, the loss of public confidence in the safe use of tunnels have necessitated the development of these guidelines. This guideline is document is intended for use to identify the categorisation of road tunnels and propose methods for the protection of the structural elements to allow users to safely evacuate, to allow rescue personnel to enter the scene and effectively perform their re- quired duties and to limit damage to the tunnel. Improved specifications for tunnel fire resis- tance are required in order to mitigate the consequences of a serious fire, which could result in structural failure or complete collapse. Assessment methods are constantly being developed to demonstrate the ability of materials and insulation coatings to prevent concrete spalling and steel and metal elements from heating and melting due to rapid heating under fire exposure conditions and to mitigate both structural and economic consequences of fire. 1.3 Document Authors This document has been developed by the members of ITA working Group 6 with the assistance of individuals from PIARC Technical Committee on Road Tunnel Operation (C 3.3) and its Working Group 6, in particular Mr. Arthur Bendelious (Animateur 1999 – Present) of the USA, and Mr. Didier LaCroix (Animateur 1991 – 1999) of France. Prof. Alfred Haack of Germany has been the liaison between PIARC and this Working Group. The primary contributors for this document are: Mr. Henry A. Russell USA (Animateur/Editor 2002-Present) Mr. Jim Richards South Africa (Animateur 1997- 2002) Mr. Martin Muncke Germany (Vice Animateur 2003 – Present) Mr. Richard Machon Germany (Vice Animateur 1997- 2003) Prof. Alfred Haack Germany (Liaison with PIARC, ITA Tutor) Mr. Ian Barry United Kingdom Mr. René van den Bosch Netherlands Mr. Per Boman Sweden Mr. Gilles Hamaide France Mr. Andreas Henke Switzerland Mr. Niels P. Høj Denmark Mr. Didier LaCroix France Mr. Yoshikazu Ota Japan Mr. Gyorgy Posgay Hungary Mr. Volker Wetzig Switzerland Dr. Kees Both Netherlands The primary contributors wish to express their appreciation to the many members of the Work- ing Group, who have provided stimulating debate and comments on this document. 03/05/2005 2-1 2. Design Criteria for Fire Resistance 2.0 Introduction The design of tunnels to be resistant to damage as a result of vehicle fire is a topic that has be- come an important design parameter for the construction of new tunnels and the rehabilitation of existing underground facilities. The objectives of tunnel structural fire resistance are to allow for users to evacuate safely in the event of a tunnel fire, rescue operations to be performed under safe conditions, and tunnel to suffer minimal loss of property. The primary focus of this document is to provide for the safety of the public and rescue person- nel in the event of a fire within the tunnel system. For these purposes, the tunnel system must be protected from collapse during a specified time period. The potential for collapse is particularly important for submerged tunnels (Immersed Tube Tunnels), and tunnels in urban environment that are located under other buildings or structures. The second focus of this document is to en- sure appropriate protection of property, which also involves indirect costs associated with the disruption to business, the local economy and the restoration of the facility to normal operation. The World Road Association, (PIARC) has issued two reports related to tunnel fires. Both were prepared by PIARC Working Group 6 Fire and Smoke Control and edited by the PIARC Tech- nical Committee on Road Tunnels (now Technical Committee on Road Tunnel Operation – C3.3): − The report Fire and Smoke Control in Road Tunnels [2.1] was published in 1999. This document is the results of the review of numerous fire scenarios and actual case histo- ries. The report has identified typical scenarios for heat generation caused by a fire event. − The new report Systems and Equipment for Fire and Smoke Control in Road Tunnels [2.3] was published in 2004. It includes a final recommendation on the question of de- sign criteria for resistance to fire for road tunnel structures. 2.1 Data on Tunnel Fires from the PIARC Report of 1999 In Section II.4.1 of the PIARC report of 1999 [2.1] indicates the heat release from vehicle fires is dependent on many variable factors such as: − Number of vehicles involved − Type of vehicles (passenger cars, coaches, heavy goods vehicles, petrol tankers) − Cross section of the tunnel − The type and quantity of flammable material available − Rate and method of extinguishing the fire The temperature is highest on exposed surfaces and particularly at the higher elevations of the tunnel structure. This is illustrated in Figure 2.1 03/05/2005 2-2 Figure 2.1 PIARC Maximum Temperatures within Tunnel Cross Section [2.1] PIARC has also established from tests the limits of the areas affected by a typical tunnel fire. These limits were based on a single fire event occurring in a tunnel. This information is useful structurally in the determination of areas to be inspected for damage. However, since the loca- tion of a fire within a tunnel is random and may occur at any location, the entire tunnel must be designed to resist fire. Figure 2.4 documents the maximum ceiling temperature in relation to the fire location for various scenarios as found in the Eureka tests. Figure 2.1 indicates the temperature distribution within the test tunnel. There seems to be little difference between the temperature distributions at the ceiling and near the road surface, when the cause of the fire is Heavy Goods Vehicles (HGV) and public buses. In the cases of plastic cars and private cars, the temperature is relatively low at the road surface when compared with that at the ceiling. The reason for this is that HGV and buses have a relatively large cross-sectional area, or in other words the ratio between projected area of the vehicle and tunnel cross-section (blockage ratio), which is relatively high. The blockage ratios between vehicle and cross sectional area of actual road tunnels are relatively lower than those for the fire tests. This means that the measured val- ues in Figure 2.1 include some allowance of safety for the correspondence to tunnel safety plan- ning. Even if the conditions pertaining to fire source in the tunnel are similar, temperature dis- tribution is influenced by the tunnel configuration. Figure 2.2 PIARC Maximum temperatures in the ceiling area of the tunnel [2.1] In addition to the maximum temperature and limits of various fire scenarios, the report also pre- sent temperature vs. time duration plots for the various types of fires encountered in the Eureka Tests as shown in Figure 2.3 and the full scale fire tests carried out in the Runehamar Tunnel in Norway in September 2003 in figure 2.4. Temperatures recorded in actual tunnel fires such as the Channel Tunnel, Mont Blanc and St. Gotthard should also be considered. 03/05/2005 2-3 Taking all current test data into account, this document considers On the whole, the 1999 PIARC report considered that the following maximum temperatures at the tunnel wall or ceiling could be developed [2.1] for the following vehicles types: • Passenger car 400° C* • Bus/small lorry 700° C* • Heavy Lorry (HGV) with combustible goods 1,350ºC (not petrol or specially flammable materials) • Petrol Tanker (general case) 1,350ºC • Petrol Tanker (extreme case) 1,400° C (for instance to avoid flooding of an immersed tunnel with bad drainage) * Higher if flames touch the wall Figure 2.3 Time Dependence Temperature Data from Eureka Test Program [2.1] Truck Figure 2.4 Heat Release Rate Curves from the Runehamartunnel Tests 2.2 Design Criteria Established by PIARC After the publication of its 19999 report, the PIARC Working Group 6 Fire and Smoke Control 03/05/2005 2-4 continued its work on resistance to fire of road tunnel structures. This paragraph presents its final recommendation on the question of design criteria for resistance to fire for road tunnel structures, which has been included in the 2004 PIARC report System and Equipment for Fire and Smoke Control in Road Tunnels [2.3] Although fire safety engineering is more and more used in other fields as a performance-based approach to fire resistance, it had been agreed between PIARC and ITA that such an approach was currently not mature enough with regard to tunnels. Consequently the joint PIARC-ITA effort has aimed at a deterministic temperature-time curve design (more classical, prescriptive approach), at least as a first step. A first draft proposal was put forward to ITA in February 2001 as an initial definition of the resistance objectives. The revised recommendation below takes into account the comments and contributions to the discussion from ITA. The first draft proposal of February 2001 was also presented to the International Standardisation Organisation (ISO) and the European Committee for Standardisation (CEN). Their comments have been taken into account in this revised recommendation, especially as regards to future steps. Comments from the members of the PIARC Technical Committee on Road Tunnel Op- eration were also integrated. 2.2.1 PIARC Recommendation A preliminary and basic criterion to be met by any tunnel structure is that there should not be any risk of progressive collapse: the local failure of any element should not lead to an increased load on other parts of the structure which may cause their failure. − Terms of Reference for the Time-Temperature Curves There are several time-temperature curves proposed to the date. Figure 2.4 sketches the ISO 834, RWS, RABT (former ZTV) and a modified Hydrocarbon (HC) curve, HC , in which inc the temperature are multiplied by a factor of 1300/1100 from the basic HC curve of Eurocode 1 Part 2-2. Figure 2.4 PIARC Recommended Temperature versus Time Curves for ZTV, RWS, HC , and inc ISO Standards[2.2] 03/05/2005 2-5 Recommendations for design of the structure should consider the time-temperature curve with regard to the possible events within the tunnel. Hence the early stages of the fire development, following the first part of the curve, will require a consideration of escape and the time con- ceived for evacuation. There should be no collapse during this period that can affect the zones where there may be users or rescuers. Spalling of the structure can occur in the early stages of a fire but no incidents have been re- ported where it has had major consequences for firemen, although it may indicate a rapid dete- rioration of the structure. The main concern at the time of fire service intervention would be the collapse of items, such as jet fans, signs or lights from the tunnel ceiling or walls. This question of fire resistance has been addressed in the PIARC 19999 report [2.1], which states: “In all cases, the minimum requirement is that heavy equipment should not fall down when evacuating users or rescue personnel are in the tunnel. This means no heavy item must fall un- der exposure to temperatures of 400-450°C during the time necessary to fight fire (in a tunnel, such temperatures can produce a radiation level of about 5 kW/m2, which is the maximum toler- able value for firemen).” Shelters should only be provided in a tunnel if there is an escape way for rescuers to reach the users waiting in the shelter and assist them to the outside. If such shelters are available, then a resistance of about two hours would be needed for protection prior to rescue. The overall duration defined by the curve will need to be considered. For instance, in France current thinking is 2 hours for the fire brigade intervention; after 2 hours it would be considered to be unsafe. If the tunnel is under a building and in other cases where protection of property is an important issue, then a longer time may be considered. − Proposed Guidelines The proposed guidelines for design criteria are presented in Table 2.1. This table makes a dis- tinction according to the type of traffic (consequently the possible fire load) and the conse- quences of a structural failure due to a fire (when the consequences are unacceptable, a protec- tion against a very severe fire is required – e.g. submerged tunnel or in unstable ground; when the consequences are limited, no protection is needed – e.g. tunnel in stable ground. Main Structure Secondary Structures4 Immersed or Tunnel Emer- Traffic Tunnel Cut Emergency under/inside in sta- Air gency Type in unstable & exits to Shelters6 superstruc- ble Ducts5 exits to ground Cover other tube ture ground open air Cars/ ISO ISO ISO ISO ISO ISO 2 2 Vans 60 min 60 min 60 min 30 min 60 min 60 min Trucks/ RWS/ ISO RWS/ RWS/ RWS/ HC ISO Tankers inc HC 3 3 120 HC HC 120 min1 inc 30 min inc inc 120 min1 min 120 min 120 min7 Table 2.1: Recommendations of PIARC 1 180 min maybe required for very heavy traffic of trucks carrying combustible goods 2 Safety is not a criteria and does not require any fire resistance (other than avoiding progres- sive collapse). Taking into account other objectives may lead to the following requirements: − ISO 60 min in most cases − No protection at all if structural protection would be too expensive compared to cost and 03/05/2005 2-6 inconvenience of repair works after a fire (e.g. light cover for noise protection) 3 Safety is not a criteria and does not require any fire resistance (other than avoiding progres- sive collapse). Taking into account other objectives may lead to the following requirements: − RWS/HC 120 min if strong protection is required because of property (e.g. tunnel under inc a building) or large influence on road network − ISO 120 min in most cases, when this provides a reasonably cheap protection to limit damage to property − No protection at all if structural protection would be too expensive compared to cost and inconvenience of repair works after a fire (e.g. light cover for noise protection) 4 Other secondary structures should be defined on a project basis 5 In case of transverse ventilation 6 Shelters should be connected to the open air 7 A longer time may be used if there is a very heavy traffic of trucks carrying combustible goods and the evacuation from the shelters is not possible within 120 min Table 2.1 uses the ISO curve and either the RWS or the HC curve to define design criteria for inc different circumstances. PIARC believes that the RWS and HC curves correspond to very inc similar levels of fire resistance, and only one of the two should be used. ISO TC92/SC2 also believes that which one you use has no impact, but considers that the HC curve is a more inc natural, better choice, should one only be kept. Currently PIARC proposes that any of these curves can be used, with very similar results. 2.2.2 Future Steps − Introduction of Tunnel Fire Curves into European and International Standards PIARC has contacted the European Committee for Standardisation (CEN/TC250 ”Structural Eurocodes”) and proposed that a temperature-time curve representative of very severe tunnel fires (either RWS or HC ) be introduced into the relevant European standard. CEN/TC250 an- inc swered in March 2001 that there was no fundamental objection to the inclusion of such a new curve. However, the Eurocode dealing with “Actions in case of fire” was in the process of being converted from a pre-standard into a full standard, and it was too late to include any new mate- rial. This should be considered at the first revision of the Eurocode. At the same time, the intro- duction of the supporting calculation rules should be considered for inclusion in the “material” dependent Eurocodes. In the meanwhile, they suggested that PIARC ask CEN/TC127, in charge of fire test methods, if it would be possible to define a tunnel fire curve for fire resistance tests. This could give a more official status to tests carried out using this curve. PIARC similarly proposed ISO/TC92/SC4 “Fire Safety Engineering” to include the same tem- perature-time curve representative of very severe tunnel fires into the relevant international standard. In March 2001, ISO/TC92/SC4 answered that they were hesitant to recommend only one temperature-time curve as being representative of fires that could develop in all tunnels. Indeed their mandate was to recommend how design fire scenarios and design fires should be tailored for a fire safety engineering assessment of a specific facility. A similar answer was re- ceived in August 2001 from their colleagues of ISO/TC92/SC2 “Fire Containment”. This issue is dealt with in the following paragraph. − Fire Safety Engineering Both ISO/TC92/SC2 and SC4, as well as several members of PIARC Technical Committee on Road Tunnel Operation and ITA WG6, mentioned that every tunnel is unique as concerns the development of a fire scenario and numerous parameters are of importance: − The type and density of traffic, and consequently the fire load and its distribution (area), as well as the possible fire spread 03/05/2005