The safety of tunnel users is the tunnel operator’s raison d’etre. The majority of papers addressing the use of active fire suppression in road tunnels review the improvements to user safety that a properly engineered active fire suppression system can bring. However, this paper will explore some of the wider issues in the specification, ownership and use of these systems. The knowledge of many years, arising from experiences of many different tunnel incidents, has impressed upon tunnel operators the seriousness of the structural damage likely to be caused to a road tunnel by a large tunnel fire. Tunnel fires can generate large quantities of very hot (up to 1000°C) smoke. This smoke is thermally buoyant and tends to accumulate in a hot layer immediately below the soffit of the tunnel.

Passive Practice
In the UK, as in many other European countries, it has become standard practice to include the installation of passive fire protection as a part of the design considerations for a tunnel refurbishment. The application of passive fire protection serves to thermally insulate the structure of the tunnel from the effects of a fire, thereby protect it from fire damage. The application of thermal insulation does not, of course, reduce the amount of heat energy emitted by the fire. It simply prevents that heat energy from being absorbed by the tunnel structure. It follows that the heat energy from a fire within a thermally insulated tunnel must, substantially, propagate within the bore. The overall effect must be an increase in temperature within the bore.

Commonly cited design specifications are that under fire exposure, the surface of the tunnel concrete should not exceed 380ºC and the reinforcement temperature should not exceed 250ºC. These factors are used to help determine the design for the passive fire protection.

But what are the equivalent heat exposure factors for the tunnel Life Safety Engineering Systems (LSES)?

It is worth defining what is meant by LSES and the use of this term. LSES is a collective term for what many people refer to as ‘the tunnel M&E’ and ‘the tunnel ITS’. The author uses LSES because Life Safety Engineering System actually tells you what the system does for you – very much in the same way that the terms ‘Bridge’, ‘Vehicle Restraint System’ or ‘Pavement’ do. As an added bonus it can also make accountants think twice before they strike a red line through the requirements to install or replace an essential tunnel safety system.

Cabling concern
All tunnel LSES contain wires and/or cables – maybe a good starting point to start to look at potential fire damage is to examine in more detail the structure of tunnel cables. Tunnels contain a lot of cables: lighting cables, ventilation fan cables, CCTV power and video cables, instrumentation cables, communications cables and so on. The traditional UK specification for these cables is that they should be ‘Low Smoke Zero Halogen’ – they do not emit toxic gases in a fire.

Cables traditionally used in UK tunnels were manufactured to BS 6724. This is not a fire-resistant cable standard (a fire-resistant cable is defined as a cable that is designed to continue to function whilst exposed to fire). There are still many thousands of miles of such cable in use in UK tunnels (at opening, the A55 Conwy tunnel contained 240km of BS 6724 lighting cables). For newer tunnels, or tunnels that have undergone major refurbishment, it has become common to install BS 7846 cables within the bore as they are likely to be exposed to fire. In construction, BS 7846 cables are very similar to BS 6724 cables; they are manufactured using Low Smoke Zero Halogen materials though they also include a mica tape inside the cable that binds the inner cores together. This tape ensures that the cable maintains its structure (and hence continues to function) when exposed to fire. Following fire exposure, such cables are also unfit for continued service – just like BS 6724 cable, BS 7846 cables also have to be replaced.

Relevant approval
The British Approvals Service for Cables (BASEC) provides a few very relevant facts regarding electrical cables. Cable life expectancy is a function of cable operating temperature. At 70ºC the design life of an electric cable is 20 years. If the same cable is operated at 80oC the design life reduces to seven years. At 100ºC the design life is "a matter of months". Cable life expectancy is highly dependent upon operating temperature. For most road tunnel applications the electrical distribution design results in the cables being operated far below the 70ºC design temperature, with 30ºC to 40ºC the much more normal operating temperatures, with in the cable life sensibly extending to 40 years or more.

Another important issue is the materials used to manufacture the insulation for be "Low Smoke Zero Halogen" cables. In the British Insulated Callender’s Cables (BICC)’s ‘Electric Cables Handbook’ by Professor G. F. Moore, we are advised that "Low Smoke Zero Halogen" cables insulation filler material includes:

  • Alumina Trihydrate – this degrades at +200ºC and forms Water
  • Magnesium Hydroxide – this degrades at +300ºC and also forms water

Damage assessment
Discussions with cable manufacturers indicate that there is no authenticated method of assessing the reduction in residual life of a cable that has been subjected to fire, or high temperature, exposure. Whilst the cable might continue to function (electrical pressure testing would indicate that the cable is ok), the life expectancy of the cable would have been significantly reduced. Cable manufacturers advise that such cables are replaced. From personal experience, the author has witnessed the testing of two cables installed immediately adjacent to each other within the same part of a tunnel fire zone. One cable was on fault (phase – earth) the other tested ok – normal phase – earth insulation levels were recorded.

To summarise the above; tunnel fires generate significant quantities of very hot smoke that accumulates around the tunnel soffit and envelopes the tunnel LSES. Passively fire protecting a tunnel will result in an increase in temperature in the bore during a tunnel fire as compared to an equivalent tunnel without passive fire protection. In addition to the tunnel cable infrastructure, tunnel LSES contains cables. All of these cables are temperature sensitive and degrade when exposed to temperatures above ~200ºC. If you have a reasonable fire, for example a ~20MW coach fire in your tunnel, there could be very significant and potentially unquantifiable damage to the tunnel LSESs exposed to the hot smoke.

In the UK if ‘street furniture’ is damaged by a vehicle it is common practice to recover the damages from the vehicle’s insurer. If we are unable to quantify the damage caused by a fire, how are we able to recover the costs? Indeed, how can we even prove that the damage (specifically the reduction in residual life) has occurred?

Clearly, we need to ensure that the information we are able to obtain about the heat exposure to which our tunnel LSES has been exposed is quantified. If (for example) we were to install thermocouples at the tunnel soffit (say) every 25 to 50m and log the data from them as a part of our normal SCADA logging systems then we would have some heat exposure data and be in a position to substantiate our damages claims. Similarly, it might be possible to obtain such data from suitably configured linear heat detection system. The method of data collection isn’t really the important issue; it’s having the data such that we can substantiate our insurance claims.

Changing policy
The attitude of insurers is also starting to become very interesting. UK insurers appear to be taking the view that damage to tunnel LSES is now a foreseeable consequence of a tunnel fire and the tunnel owner ought to take steps to mitigate such damage. For a modern tunnel, the tunnel LSES costs can amount to many millions of dollars. The post incident recovery time can also be very significant. It was estimated that the long-term closure of the Mont Blanc tunnel caused some EUR 2.5bn (USD 3.22bn) in regional economic damage.

Rewiring our existing tunnels with fire-resistant cables will help to ensure that essential LSES stand more of a chance to continue to function during a tunnel fire – but it would still have to be replaced after a fire.

Given that in a road tunnel a BS 6724 cable installation could have in excess of a 40-year life expectancy, it seems poor value to replace it half way through its life with a BS 7846 cable that is equally as susceptible to being written off through fire damage (despite being able to continue to function during a larger fire). We could also consider enclosing the cables within a thermally protected enclosure. The specification for this would have to be to ensure that the cables inside the enclosure did not exceed 200oC as a maximum (100oC would provide a more comfortable factor of safety). Such an enclosure is likely to be significantly thicker than the equivalent thermal protective system used for concrete, as we only need to limit the concrete surface temperature to 380oC. Enclosing cables within a thermally protective enclosure prevents them from dissipating heat to the environment during their normal operation. Such a measure is likely to result in the necessity to de-rate the cables (perhaps to 1/3 of their previous rating), so we’d probably have to replace them in order to be able to do this – for a large number of existing tunnels this could be a pointless exercise.

An alternative approach might well be to install active fire suppression.

Active fire suppression systems are usually designed to control fire growth. It is usually the case that a tunnel fire starts small and grows, sometimes very quickly, into a large fire. Timely activation of a tunnel’s active fire suppression system can limit fire growth to an extent that the fire never becomes a large fire, that the high temperatures associated with a large fire are not generated and the damage to the tunnel LSES is significantly reduced. Similarly, it is also possible to design the systems to provide active cooling of the tunnel LSES and structure. This technique has been used for many years in HV distribution systems.

From this perspective, active fire suppression starts to look like a very attractive mechanism for limiting damage to LSES – this is in addition to the well-established tunnel user safety considerations. Active fire suppression seems to have the potential to be both a tunnel life safety system and an asset protection system.

Types of active fire suppression
There are numerous different types of active fire suppression systems; mainly either water based or foam based. A full review of all of them is beyond the scope of this paper. For the purpose of this paper, the consideration is restricted to the salient aspects of the two main types of system commonly in use throughout the world. High-pressure ‘mist’ systems are systems that generate a fine mist, while low pressure ‘deluge’ systems are systems that produce an intense water spray.

The main characteristics of a high-pressure mist system:

  • It generates a fine mist that is an effective block to radiant heat and produces significant visual obscuration
  • Fine mist is more susceptible to tunnel ambient or ventilation induced air flow than water spray owing to the much smaller water droplet size and correspondingly lower droplet momentum from a mist
  • It uses less water (1/3) than a deluge system to create the same fire suppression capacity
  • It uses high pressure water distribution of ~140 to 200 bars throughout the tunnel
  • It uses a large number of high pressure pumps that require a high electrical supply capacity (approximately 1 – 1.5MVA / installation) or large prime movers
  • It requires the supply water to be finely filtered
  • System maintenance and support is relatively specialist necessitating specialist support

The main characteristics of a low-pressure deluge system:

  • It generates a water spray that is a less effective block to radiant heat than fine mist – the deluge is also likely to produce visual obscuration
  • Water Spray is less susceptible to tunnel ambient or ventilation air flow than fine mist owing to the much bigger water droplet size and correspondingly higher droplet momentum from a spray
  • It uses approximately three times as much water to create the same fire suppression capacity as a deluge system
  • It uses low pressure water distribution of around 5 to 10 bars throughout the tunnel. Zone valves and so on also operate at a low pressure
  • It uses a few low pressure high volume pumps that do not require a high electrical supply capacity, at around 0.35 to 0.55MVA / installation
  • It does not require the supply water to be finely filtered
  • ystem maintenance and support is relatively commoditised and more widely supportable

From the above it quickly becomes apparent that the retro-fitment of a mist system may necessitate extensive increases in the tunnel electrical network and supply capacity or large prime movers, and the associated increase in electricity supply costs. It also appears reasonable to expect that the maintenance and support costs would be comparatively high.

The flooding risk appears lower than a deluge system.

Similarly, the retro-fitment of a deluge system can be expected to necessitate modest increases in the tunnel electrical network and supply capacity or smaller prime movers, and a comparatively smaller increase in the associated electricity supply costs. It also appears reasonable to expect that the maintenance and support costs would be comparatively low, though the flooding risk would be higher.

System selection, installation and operational issues
Designers of both types of system appear to standardise the basic operation of each system type on the simultaneous activation of three zones – a zone typically being 25 to 30m in length.

A tunnel operator knows that there is a much higher probability of experiencing a car fire than a lorry fire. Following an in-depth study for one tunnel, the projected ratio was 43 :1 (i.e. 43 car fires for every 20 to 30MW lorry fire). To put another way, 97.5 per cent of these fires will be small car fires. This really begs the question of why an operator needs to flood 75 to 90m of tunnel to control a small car fire.

Defining the activation zones for an active fire suppression system to be, for example, 10 to 12m would enable the tunnel operators to activate the AFS in accordance with the size of the fire.

A further consideration is that of inadvertent system operation (a zone ‘goes off’ during normal traffic flow). At 50mph (80kph), a 10m activation zone would cause visual obscuration for less than 0.5s travel time.

A 30m activation zone would cause obscuration for nearly 1.5s travel time. The quantity of water used by a deluge system would be more likely to cause localised carriageway flooding and the possibility of aquaplaning than would a mist system. A reduction in activation zone length (i.e. increasing the number of activation zones) appears to have some considerable merit from the perspective of tunnel operations.

Moving to improve
A key question for tunnel operators is how to operate these systems. The Road Tunnel Operators Association (RTOA) has recently, in conjunction with the Awarding Body ProQual, issued a Qualifications and Credit Framework (QCF) Level 3 Diploma in Road Tunnel Operations.

With an eye to the future, this Diploma contains a module entitled ‘Operating Tunnel Life Safety Engineering Systems – Active Fire Suppression’. This module can be taken either as a part of the Level 3 Diploma, or as a part of a Level 3 Certificate or on its own as a stand-alone module.

This mechanism allows a tunnel operator to configure an approved training package for the systems they intend to install – much of the technical training content for the system could be procured as a part of the AFS system procurement.

Of vital importance is the occupational competence of persons involved in the provision of training. It is the employer’s legal responsibility to ensure that their employees (including external training service providers) are competent.

The RTOA has tortured itself getting to grips with this issue.

It has now established an Occupational Competence Validation Committee (OCVC), headed by a senior industry operation practitioner. The function of this committee is to assess the occupational competence of individuals involved in the tunnel operational training industry, much in the same way that we use prequalification procedures to ensure that tenderers for works are competent. Initially, we expect this to be focussed on training assessors, though there is, of course, an opportunity for the scope of this to be expanded should the Association consider this to be desirable.

The Diploma in Road Tunnel Operations should now provide tunnel LSES designers with a firm basis on which to agree the necessary Operator input to tunnel system designs such that the required level of overall system safety is achieved.

Final Remarks
Whilst the industry does currently appear to be tending to the use of mist systems, the author would argue that for many applications a deluge system may be a better proposition, as it is likely to be cheaper to retro-fit, to own and also to maintain.

Being able to apply a significant volume of water to a fire has a certain quality about it – as Joseph Stalin once observed – "Quantity has a quality all of its own"