TUNNEL SAFETY EXPERTS from around theworld should welcome a decision of theAustralian courts on 30 January 2013 torelease the Courts detailed technical findingsabout the sequence of events, systemsperformance and human behaviour arisingfrom the 2007 fatal Burnley Tunnel fires. Boththe Court's technical documentation and theCourt's determination provide importantimpartial information on how a modern roadtunnel performs in severely challengedemergency conditions.

Importantly these Court reports exploredriver behaviour before the crashes, thedegradation of the computer systemscontrolling the electromechanical systemsduring the event, the behaviour of survivors asthey self rescue, the effectiveness of the smokemanagement systems, the operation of thedeluge systems, the performance ofemergency services and the subsequentreopening of the tunnel.

Both the report by the Court appointedexpert and the findings and recommendationsof the Court itself provide useful insights intoan actual tunnel incident, a rare windowcreated by a legal investigation providing auseful vision of tunnel safety and a list ofrecommendations for contemporary tunnelsafety.

The Court made recommendationstouching upon education of drivers into tunnelsafety, design imperatives such as reducing theneed for lane changes, the importance ofconsistent vehicle speed within tunnels,emergency stopping lanes, the need foreffective emergency egress pathways, ensuringfunctionality of control computer systems,design aspects of deluge systems, theimportance of emergency evacuationmessages and the functionality of emergencyservices communications underground.

The Court noted with approval theobservations of the Victorian Assistant ChiefFire Officer that:

"... if it wasn't for the CityLink operatorsacting as they did there would have been afar greater catastrophe ... there would havebeen many more fatalities ... potentially

everyone inside the tunnel that day couldhave perished ..."

While this short summary is no substitute forreading the Court judgements and expertreport it provides an insight into the incidentand the basis for the recommendations madeby the Court.

The TunnelThe Burnley tunnel is one of two nonidentical pair of tunnels, three lanes wideservicing Melbourne’s busiest toll roadnetworks. Vehicle numbers typically exceed100,000 vehicles per day per tube, and thereis a high proportion of HGVs.

The Burnley tunnel is 3.4km long and at itsdeepest point is 60m under ground level. Thegrades within the Burnley Tunnel are veryhigh exceeding a 6% down slope after theentrance portal prior to its deepest point andexceeding a 5% up slope for nearly onekilometre prior to the exit portal.

The Burnley Tunnel has three, 3.5m lanes, avertical traffic clearance of 4.9m and two0.5m shoulders as well as an elevated 0.8mwide walkway. It carries a mixed fleet with asignificant proportion of heavy goodsvehicles. On the morning of the accidenttraffic mix was 28% heavy goods vehiclesand 72% cars.

Traffic flow is unidirectional; there is noprohibition on lane changes or any restrictionon vehicle types in lanes at the time of thecollisions. Dangerous goods are not permittedthrough the tunnel.

Both tunnels are longitudinally ventilated,equipped with a deluge system,supplemented by a dedicated overheadsmoke extraction system, and monitoredcontinuously by both human operators andartificial intelligence technologies. The delugesystems are part of a sophisticated computerbased safety system but are ultimatelymanually operated.

This deluge system is divided intoindividually operable zones of 30m in lengthand 11.5m in width. It is a dry system – onceremotely operated quick action valves arecommanded to open and an otherwise dryset of pipes is filled with water.

The deluge discharge rate is around 2,850litres per minute per 30m zone. The tunnelsare equipped with large collection sumps.

36 TUNNELLING JOURNAL

Professor Arnold Dix,Scientist and Lawyerexplains the recentcourt findings intoMelbourne’s fatal BurnleyTunnel fires in 2007 and whatthey mean to the tunnelling industry

FIRE SAFETY

The Burnley TunnelFires of 2007 - a Courtsperspective

Pic 1: Spiral deluge nozzle in area of thetunnel fires

TJ_0313_ARNOLD_036_039.qxd:Feature 1/3/13 20:32 Page 36

FIRE SAFETY

TUNNELLING JOURNAL 37

The IncidentAt 09:52:30 am on the 23rd of March2007 in a steep downhill section of theBurnley Tunnel, a truck made anunscheduled stop.

Over the next two minutes 103 vehiclespassed the stopped truck without incident,although traffic become congested aroundthe stopped vehicle.

Two minutes later, by 09:54:24 severalvehicles, including 4 HGVs and 7 lightvehicles had crashed, 3 people were deadand fire and a series of explosions wereinitiated.

By 09:56:00am (two minutes afterignition) emergency ventilation and a fixedfire suppression system had been activated,and an effective emergency evacuation ofseveral hundred tunnel users hadcommenced.

Fire growth and spread were contained.

Pic 2: This image shows the stoppedtruck and the red arrow shows theartificial intelligence detecting thestopped vehicle and raising an alarmfor the operator.

Pic 3: The incident unfolding

Pic 4: 9:54:27 - Controllers alarms attime of first collision, the fireball isgenerated by fuel released from thefirst crushed vehicle.

Pic 5: 9:54:42 - 15 seconds after initialimpact showing truck driver runningfrom impact scene

Pic 6: 9:55:59 - this photograph shows atruck driver after the crash protectinghimself from the radiation coming fromthe burning vehicles before the delugesystem was operated.

Pic 7: 9:56:49 - Deluge activating

Pic 8: 9:57:28 Deluge suppressingfires

Cars that become crushed

Skidding truck - tyre smoke

Truck 1 (stopped)

TJ_0313_ARNOLD_036_039.qxd:Feature 1/3/13 20:32 Page 37

Emergency services were able toapproach the incident and extinguish thefires. Tunnel damage was minimal andtolled operations could have re-recommenced within hours.

The Court summarised the sequence ofevents as follows:

"Sixty seven seconds after that firecommenced, the Tunnel operator enabledemergency mode in preparation for thesmoke extraction, deluge operation andevacuation. Twenty seven seconds later theemergency response plan was activated bythe Tunnel operator and this activated theemergency smoke extraction and thedeluge system."

CommentaryDespite the severity of these initiatingevents the fires were contained, with noflash over or other significant fire growthoccurring once the deluge fixed fire fightingsystem was initiated.

However, it was not merely the presenceof the fixed fire fighting system which wascritical – it was that the ventilation systemwas effective in that it stopped backlayering(up a steep tunnel grade of 6%) andreduced the longitudinal airflow rapidly (toapproximately 2m/sec) in order to optimizesmoke extraction and minimize ventilationinduced fire growth. It was the fire brigadethat put the fires out – the deluge systemmerely kept the fires small enough to allow

effective emergency services intervention.The incident resulted in several hundred

people being evacuated from the tunnel –and their vehicles. None of the evacuees ortheir vehicles was injured or damaged. Thetunnel only suffered minor damage, andcould have been re-opened 10’s of hourslater if the extent of the damage could havebeen more rapidly determined.

Picture 10 depicts the minimal damage tothe tunnel. Spalling of the non structuralroad barrier was confined to an area ofapproximately 100mm by 150mm.

DiscussionIn the Burnley incident the ventilationsystem rapidly reduced the longitudinalvelocity to approximately 2m/sec. This lowventilation rate was sufficient to stopbacklayering despite the buoyancy effectcaused by the tunnels steep grade of inexcess of 6.2% at the incident location.

Downwind of the incident conditionsrapidly deteriorated. Video images suggestcomplete destratification occurred duringthe post fire turbulent flow period. Smokeextraction systems coupled with inducedlow longitudinal air flows rapidly controlledthese degraded conditions.

The rapid activation of the FFFS andquick control of longitudinal velocitiescontributed to minimal fire growthfollowing the crash, explosions andsubsequent fire.

The fire inside the structure of the primemover was not extinguished by the FFFSbut was extinguished by the interventionof fire fighters. This is entirely consistentwith the expectations derived fromexperiments involving shielded fires.

The absence of flashover and lack ofaccelerated fire growth is consistent withthe experimental data on the effects of a

FFFS with water application rates roughlyin the order of 10mm/min.

Human BehaviourThe majority of people did not use theemergency cross passages or the elevatedemergency pathways to evacuate. Instead,most people used the roadway.

Many evacuees chose not to useevacuation passages - indeed several peopleactively rejected using cross passages afterbriefly inspecting them.

However those tunnel users that did usethe cross passages to access the adjoiningsafe tunnel chose to use the elevatedwalkways to self rescue.

Despite the presence of emergencytelephones no evacuees used them.Evacuees were observed actively helping theless able to escape, while disabled peoplewere observed to be amongst the fastest tomake and commence evacuation followingthe incident.

Pic 9 - 10:05:00 Fireman withoutbreathing apparatus observed to enterfire zone and fight fire undercover ofdeluge operation

Pic 10: Minimal damage to the tunnel

Pic 11: Burnley Tunnel after fire - road surface and tunnel walls and services still intact.

Pic 12: 10:05:31 - Domain Tunnel (safetunnel) evacuees that used the crosspassages to escape the fire chose to usethe elevated walkways within the safetunnel.

FIRE SAFETY

38 TUNNELLING JOURNAL

TJ_0313_ARNOLD_036_039.qxd:Feature 1/3/13 20:32 Page 38

TUNNELLING JOURNAL 39

A significant number of people chose tostay and observe the incident, many takingphotographs and putting themselves at riskdoing so.

These observations (amongst others) ledboth the Court and the Court's special expertto recommend that higher priority be placedupon the effectiveness of measures designedto protect tunnel users during the self rescuephase.

ConclusionThe rapid and accurate use of Burnley’s FFFScoupled with effective longitudinal air velocitycontrol coincided with minimal tunnel damage,no non crash human victim fire related injuriesand rapid reopening of the infrastructure.

The use of FFFS coupled with advancedtunnel ventilation control, rapid incidentdetection and accurate response positivelycontribute to tunnel fire safety and assetprotection.

The Court's RecommendationsThe Burnley incident was the subject of both aSupreme Court criminal trial and a full coronialinvestigation. The Court ultimatelyrecommended a range of measures including:• Drivers be educated about safety in tunnels,

especially with respect to safe distances, thedangers of lane changing, driver distraction,emergency behaviour, self rescue andemergency response

• Implementation of a ban on lane changingin tunnels be considered

• Tunnel designs that encourage consistencyof vehicle speed as an objective duringdetailed proposed new tunnel designassessments be favoured

• Consideration of an emergency lanerequirement for new tunnels beimplemented

• Minimisation of tunnel intersections bepursued

• Recognition of the importance of sight linesin designs be acknowledged

• An emphasis on improved performance ofemergency egress pathways for evacuees beembraced

• A focus on assessing the operationaleffectiveness of emergency controlcomputer systems be developed

• A functional review of the performance ofdeluge systems during multiple events beundertaken

• A review on the content of emergencymessages for evacuating pedestrians beundertaken

• A focus on the operability of emergencyservices communications equipment insubsurface environments become anoperational focus

These recommendations provide a pertinentreminder to us all on the importance of thesystems we create to work during emergenciesfrom a practical perspective. Theseobservations support the efforts oforganisations such as ITA COSUF, PIARC andNFPA 130 and 502 to refine and furtherimprove the requirements found in regulatoryand quasi-regulatory publicationsinternationally.

This event reminds us of the importance ofensuring the tunnels we create are functionallyresponsive to the practical needs generated bypeople whom seek to enjoy safe journeysunderground.

Pic 13 - 9:59:37 - (5 minutes after theincident) self evacuation is wellunderway with disabled, adults withpushers, and able bodied adults mostlyevacuating on foot using the roadway(some evacuees were observedreturning to their vehicles to gatherbelongings or trying to approach theincident to take photographs).

S c h a u e n b u r g M A B G m b H Wese le r S t r. 35 45478 Mü lhe im a . d . Ruh r / Ge rmany

w e b s i t e : w w w. s c h a u e n b u r g - m a b . c o m e -ma i l : hwm@hwm-eng inee r i ng .de

p h o n e : + 4 9 ( 0 ) 2 0 8 - 9 9 9 1 - 0 f a x : + 4 9 ( 0 ) 2 0 8 - 5 9 2 4 0 9

Your Reliable Partner in TunnellingSeparation Technology Worldwide 15-21 April, Munich

Booth C3.413

Visit us:

TJ_0313_ARNOLD_036_039.qxd:Feature 1/3/13 20:32 Page 39