Lisbons funicular suffers catastrophic runaway (03/09)

[neonison]

“Detached” isn’t the same as “snapped” and one may speculate that the detached portion may have also included the tension activated emergency brake (Otis style, if there indeed was such) and other parts of the braking mechanism.

The lack of an effective alternative braking mechanism would appear to be significant.

 

[bahnause]

If the above is correct then the thing is just very poorly designed, full stop. Regardless of what else was going on, I think even cable cars usually have an extra cable for redundancy no? Even if not (on googling it this doesn't seem to be a universal rule) then there is no reason this couldn't have had proper brakes.

That depends on the design. Taking the specifications for aerial ropeways as an example, an independent braking system is not required for installations with a continuous spliced cable. A continuous cable is easier to inspect. However, as soon as the cable is split and connected to the gondola, a separate braking system is required for the lift to be able to stop in the event of a defect in the cable or attachment. Defects almost always occur at the attachment rather than on the cable.

I am not sure if these regulations can be applied to funicular railways on a one-to-one basis.

EDIT: According to current EU standards, a 'safety brake' would be required in these cases. However, this standard does not apply to the Gloria Railway, as historic, culturally valuable and listed railways are partially exempt.

 

[oldman]

The initial official report in English is here. It refers to the cable 'giving way' at the attachment point to the car

This (my emphasis) caught my non-technical eye:

the evidence indicates that the air brake and the manual brake were quickly applied by the brakeman of cab #1, but that in the current configuration, the brakes do not have sufficient capacity to stop the cabins in motion without their empty masses being mutually balanced by the connecting cable.

 

[edwin_m]

If the above is correct then the thing is just very poorly designed, full stop. Regardless of what else was going on, I think even cable cars usually have an extra cable for redundancy no? Even if not (on googling it this doesn't seem to be a universal rule) then there is no reason this couldn't have had proper brakes.

I doubt it would be practical to provide a second cable linking the vehicles via the top. It would have to be led round a separate set of pulleys and probably need a second trench. And there would be difficulties with them being slightly different lengths as the vehicle passes the curves on the route, leading to one of them being slightly slack.

I believe the normal arrangement is for a brake to be held off by the tension in the cable, and it appears they had this here although pneumatic rather than spring-applied (see just after Figure 2 in the linked in #63).

In the event of a failure in the cable connection between the cabins, the funicular system is designed to:
➢ Automatically apply the pneumatic brake to each cabin at maximum force, through an internal mecha-
nism triggered by the loss of cable force in the trambolho;
➢ Cut power to the cabins' electrical system, through a tripping device located in the technical compart-
ment at the top of Calçada da Glória and incorporated into the cable's flywheel support, which detects
the absence of the load transmitted by the cable. The absence of electrical power in the system also
has the effect of automatically applying the pneumatic brake to each cabin at maximum force.

“Detached” isn’t the same as “snapped” and one may speculate that the detached portion may have also included the tension activated emergency brake (Otis style, if there indeed was such) and other parts of the braking mechanism.

The lack of an effective alternative braking mechanism would appear to be significant.

The report says the brakes were operated but insufficient to stop the vehicle without the tension in the cable.

At this time, it has not yet been possible to conduct verification checks to confirm whether the system for
automatically applying the pneumatic brake to the cabins following a loss of cable tension in the trambolho
worked.
However, regardless of this, the evidence indicates that the air brake and the manual brake were quickly applied
by the brakeman of cab #1, but that in the current configuration, the brakes do not have sufficient capacity to
stop the cabins in motion without their empty masses being mutually balanced by the connecting cable. There-
fore, the existing brakes does not constitute a redundant system in case of a failure in the connecting cable.

That's on a par with saying that a helicopter is poorly designed if it doesn't have a backup rotor to use if the primary one falls off.

That is clearly not possible, so safety of helicopters is assured by very careful design and maintenance to make it not credible for this to happen. There are other potential single point failures in transport systems, for trains most obviously the wheels and axles, which correspondingly have very close attention.

In the case of a funicular it's clearly practicable to provide a secondary mechanism able to stop the vehicle without the cable tension, as other funiculars do this. UK health and safety law would probably therefore require one to be provided as a reasonably practicable safety measure. Instead, assuming the report is accurate, they seem to be relying on the inspection regime as with the helicopter example, to justify that cable separation isn't a credible failure. This has proved to be a false assumption. Either that or they just didn't know that the brake force was insufficient to stop the vehicle, which would suggest a lack of basic understanding of the system.

 

[HighlandStorm]

On the subject of aerial cable cars there isn’t redundancy in the cables - in the sense that one is there as backup in case the other breaks. There is a track rope which carries the weight of the cabins and is equivalent to the tracks of a funicular, then a lighter weight traction rope that forms a loop around a driving bullwheel at one station and return wheel at the other.

A loss of tension in the traction rope would initiate an emergency brake application such that the cars are brought to a controlled stop on the track rope.

A failure of the track rope = game over, the traction rope is neither designed to nor has the tension to take the load carried by the track rope.

In the situation of mono-cable cableways, the function of traction and track rope are combined in a continuous haul rope, braking systems will usually consist of a regenerative motor for normal slow downs, a service brake which will usually consist of be a large disc brake connected to the bullwheel drive and an emergency bullwheel brake that clamps directly onto the bullwheel.

Something else on the funicular crash is it’s noted the power supply was cut by circuit breakers - that presumably electrically disconnected the cars so no regen either.

On whether the cable broke or it was a connection failure on the car, on fixed grip aerial cable ways it’s a requirement to periodically move the lifts carriers so grips don’t fatigue the haul rope.

A question here is whether the cable was disconnected from the cars for the inspection that had just take place?

Modern funiculars have to demonstrate they can be brought to a safe stop from an overspeed while overloaded using only the on car brakes!

 

[Bikeman78]

The report says the brakes were operated but insufficient to stop the vehicle without the tension in the cable.

Adhesion alone wouldn't be sufficient on that gradient. If the only brake force is applied to the wheels, I doubt that would even hold the cars at a stand once the cable is detached/broken, let alone stop a moving car.

EDIT: Having read the report, I see that the cars do have another braking system though I'm not entirely clear as to how it works.

 

[hexagon789]

Adhesion alone wouldn't be sufficient on that gradient. If the only brake force is applied to the wheels, I doubt that would even hold the cars at a stand once the cable is detached/broken, let alone stop a moving car.

The wheel tread brakes are not the only braking system.

The main braking is done by cast iron shoes pressing on the metal rim of the cable trench.

There is no rheostatic/regen brake.

 

[Gag Halfrunt]

The British victims have been named.

Families of British couple killed in Lisbon funicular crash ‘heartbroken’

Relatives pay tribute to Kayleigh Smith, 36, and William Nelson, 44, who died when the Elevador da Glória derailed

www.theguardian.com

The families of a British couple killed in Lisbon after a funicular streetcar derailed have paid tribute to them and have said they are “heartbroken”.

Kayleigh Smith, 36, and William Nelson, 44, died alongside 14 other people after the Elevador da Glória hurtled down a hill and careered into a building on Wednesday night.

Smith, from Macclesfield in Cheshire, graduated from the Arden School of Theatre in Manchester, where Nelson worked as a theatre director. The identity of the third British victim has yet to be disclosed by authorities.

In a statement issued by Cheshire police on Saturday, Smith’s family said: “Kayleigh was loved by family and friends for her wit and humour, her kind and caring nature came to the fore in her work as a funeral operative.

“She was also a talented theatre director and had just completed a master’s degree. They both leave family and friends heartbroken.”

In the same police statement, Nelson’s brother, who was not named, said: “Words cannot begin to describe how our family and friends are feeling right now but here is the best attempt. This week, due to a tragic accident in Lisbon, Portugal, we lost Will Nelson, who was not just my big brother but everyone’s.

“He was always kind, selfless, and protective, and the world does not feel right or normal without him. He was and has always been my hero, and we will miss you always. Love you, rest in peace you legend.”

The accident, described by Portugal’s prime minister, Luís Montenegro, as “one of the biggest tragedies in our recent history”, also left 21 people injured, five of whom are in a serious condition.

In a statement on Friday morning, police said the accident had killed five Portuguese citizens, three Britons, two South Koreans, two Canadians, one American, one Ukrainian, one Swiss citizen and one French person.

Police said a German citizen thought to have died in the crash had been located alive in a hospital. Local media had reported that a German father had died and a mother was seriously hurt while their three-year-old child suffered minor injuries.

Hours before the crash, Smith posted pictures on Instagram of the couple’s first day in Lisbon with the caption: “Churches and castles, tiles and trams.”

MADS Theatre in Macclesfield posted a tribute on Friday night saying: “Kayleigh was a valued member of our society and made considerable contributions both to MADS and to drama in the North West. It is indeed a sad loss to all of us at the theatre.

“Kayleigh was an award-winning director and an award-nominated actress. She also undertook multiple crew and front-of-house roles at MADS. She was a past vice-chair, membership secretary and head of tech, but above all that, she was a dear friend to many and will be greatly missed.

“We wish to send our deepest condolences to both families, and to respect their privacy at this sad time. Our thoughts are with them.”

André Jorge Gonçalves Marques, who worked as the brake guard on the funicular, was among the fatalities, along with Pedro Manuel Alves Trindade, a former volleyball referee.

The Portuguese transport union Sitra said: “We send our condolences to the families and friends of the victims of the accident and wish them a speedy recovery as well as the best recovery to the others injured in the accident.”

Marques was a “dedicated, kind and happy professional, always willing to contribute to the greater good”, according to his employer, Carris, which runs the funicular.

Investigators have not found any evidence of sabotage, leaving mechanical failures or maintenance issues among the possible causes.

The Glória line carries about 3 million people annually. Its two cars, each capable of carrying about 40 people, are attached to opposite ends of a haulage cable, with traction provided by electric motors on the cars.

A No 10 spokesperson said: “The prime minister is deeply saddened to hear that three British nationals have died in the funicular crash in Lisbon. His thoughts are with their families and those affected by this terrible incident. We stand united with Portugal during this difficult time.”

 

[bahnause]

I still don’t understand the organisational aspects. It's not entirely clear yet which set of rules the cable car was operated under, or who was supervising it. The statement that all necessary inspections were carried out says little about whether they were appropriate or which regulations they were derived from.

I remember the Liberec–Horní Hanychov–Ještěd aerial cableway crash. There, the same handful of employees worked in different roles: sometimes as test technicians, sometimes as maintenance technicians, operations managers or cabin crew. Sometimes they worked in different companies and sometimes in the same one. Such conflicts of interest are prohibited in other industries.

 

[Taunton]

Cable breaks are by no means unknown; there are videos on the Internet from San Francisco showing broken cables, and the museum at the powerhouse there has one on display, along with the extensive measures at the winding wheels to detect strands coming loose, which are typically predecessors to an actual break. Despite being Victorian technology itself there is no single point of failure in the system. The "hot lever" emergency brake is quite independent of the cable grip mechanism.

 

[petergb]

The main braking is done by cast iron shoes pressing on the metal rim of the cable trench

Overhead photography shows that the cable trench is all ripped up and destroyed in the section just before the crash site.
https://eu.usatoday.com/videos/news...lisbon-funicular-crash-aftermath/85974572007/

I suspect that form of braking system and the construction of the cable trench may simply be incapable of providing adequate arresting force to cope with a runaway vehicle. The brake system tried to stop the vehicle, but ended up destroying the trench instead.

(edited to correct typo)

 

[hexagon789]

Cable breaks are by no means unknown; there are videos on the Internet from San Francisco showing broken cables, and the museum at the powerhouse there has one on display, along with the extensive measures at the winding wheels to detect strands coming loose, which are typically predecessors to an actual break. Despite being Victorian technology itself there is no single point of failure in the system. The "hot lever" emergency brake is quite independent of the cable grip mechanism.

Cable breakages are not a problem in San Francisco because the track and wheel brakes are sufficient to stop and hold a car independently of the cable.

The Gloria funicular, and it appears 2 of the other 3, seem to be designed such that the cable forms an integral part of being able to stop.

This personally surprises me, as the main braking system of brake shoes pressing onto the cable trough rim seems not dissimilar in principle to a track brake pressing on the rail surface - ordinary wheel/track adhesion isn't necessary to decelerate or stop, the friction level is much greater.

== Doublepost prevention - post automatically merged: 7 Sep 2025 ==

Overhead photography shows that the cable trench is all ripped up and destroyed in the section just before the crash site.
https://eu.usatoday.com/videos/news...lisbon-funicular-crash-aftermath/85974572007/

I suspect that form of braking system and the construction of the cable trench may simply be incapable of providing adequate arresting force to cope with a runaway vehicle. The brake system tried to stop the vehicle, but ended up destroying the trench instead.

(edited to correct typo)

Hmm, it seems that way.

I appreciate that things were different in the past, but even the Victorians were usually good at ensuring some level of redundancy with transport operating in more extreme circumstances.

The use of shoes pressing on the cable trough, to me seems that with sufficient braking force applied it should be able to stop and hold a car on the gradients even without the cable balancing force.

Either the brake force provided is insufficient for this or the cable trough cannot take that strain - whether either of those is by design fault not appreciated or even whether the original design contemplated the need to stop without the cable balancing force I'm not yet clear on.

 

[edwin_m]

Adhesion alone wouldn't be sufficient on that gradient. If the only brake force is applied to the wheels, I doubt that would even hold the cars at a stand once the cable is detached/broken, let alone stop a moving car.

EDIT: Having read the report, I see that the cars do have another braking system though I'm not entirely clear as to how it works.

The wheel tread brakes are not the only braking system.

The main braking is done by cast iron shoes pressing on the metal rim of the cable trench.

There is no rheostatic/regen brake.

Do these shoes press downwards or laterally?

If downwards, the total decelerating force, whether applied by wheels or by is limited by the weight of the vehicle and the coefficient of friction. So if the percent gradient is greater than the coefficient of friction the vehicle cannot declelerate*. At an average gradient of 18% the coefficient of friction needed for any deceleration is close to the maximum achievable with steel on steel.

This wouldn't apply if the brake acts laterally on the sides of the trench, as it would then generate its own reaction force independent of the vehicle weight. It also wouldn't apply to a magnetic track brake, which if I have my physics right would create a reaction force additional to the weight, that also increases the downward force on the wheels.

From the report it travelled 170 metres. Assuming the quoted average gradient of 18% this is a vertical descent of 30m. With no braking and no other friction or other resistance losses this would give a final speed of 24m/s or 86km/h. They tentatively estimate the speed reached was 60km/h, implying that the available braking was not just not enough to stop the vehicle, it was nowhere near being enough.

*This relies on small angle assumptions which may make it only approximately true at that sort of gradient.

Overhead photography shows that the cable trench is all ripped up and destroyed in the section just before the crash site.
https://eu.usatoday.com/videos/news...lisbon-funicular-crash-aftermath/85974572007/

I suspect that form of braking system and the construction of the cable trench may simply be incapable of providing adequate arresting force to cope with a runaway vehicle. The brake system tried to stop the vehicle, but ended up destroying the trench instead.

The report says that the vehicle started to overturn, but was initially restrained by the cable grip rubbing against the underside of the trench slot, until the forces became enough for the cable grip to rip the top of the trench out of the road. This is a potential alternative explanation for why the trench is damaged just before the crash site.

 

[Egg Centric]

That's on a par with saying that a helicopter is poorly designed if it doesn't have a backup rotor to use if the primary one falls off.

I don't agree with that. A secondary rotor isn't practical. A brake that's capable of holding / stopping something on an 18% gradient on eg a rack and pinion type system clearly is.

 

[hexagon789]

Do these shoes press downwards or laterally?

If I've understood the report correctly, they seem to press outwards against the sides of the cable trough metal rim - so laterally.

 

[Paul AC]

If I've understood the report correctly, they seem to press outwards against the sides of the cable trough metal rim - so laterally.

No; vertically as I read it. https://www.gpiaaf.gov.pt/upload/processos/d054239.pdf figure 2 shows red 'braking effort' arrows acting vertically to 'squeeze' the Z-profile flanges to create the emergency braking effort. That's OK as long as the Z-flanges can withstand the forces involved. What appears to me to have happened here (speculation) is that the Z-flanges have been pulled out of the roadway by the forces involved, as 'edwin_m' and 'petergb' have said above; suggesting that they were structurally inadequate.

 

[hexagon789]

No; vertically as I read it. https://www.gpiaaf.gov.pt/upload/processos/d054239.pdf figure 2 shows red 'braking effort' arrows acting vertically to 'squeeze' the Z-profile flanges to create the emergency braking effort. That's OK as long as the Z-flanges can withstand the forces involved. What appears to me to have happened here (speculation) is that the Z-flanges have been pulled out of the roadway by the forces involved, as 'edwin_m' has suggested; suggesting that they were structurally inadequate.

Ah, my apologies.

The diagram seemed to suggest an outward, scissor-like operation of the brake calipers but presumably it's actually sprung and the whole mounting then presses downwards.

 

[bahnause]

I don't agree with that. A secondary rotor isn't practical. A brake that's capable of holding / stopping something on an 18% gradient on eg a rack and pinion type system clearly is.

I took the Cardada cable car this morning to go for a hike. It was built in 2000, has two 40-seater cabins and no catch brake. So if the cable breaks, the cabin will go straight down into the valley.

Not everything that is technically feasible or practical is implemented; as always, a risk assessment is conducted.

 

[Bletchleyite]

I took the Cardada cable car this morning to go for a hile. It was built in 2000, has two 40-seater cabins and no catch brake. So if the cable breaks, the cabin will go straight down into the valley.

Not everything that is technically feasible or practical is implemented; as always, a risk assessment is conducted.

I have always wondered if this was the case particularly on older funiculars. If it indeed is, I will now be much more reluctant to use them; this to me is a serious design flaw and places an unacceptable risk of failure on one component, where if a failure occurred deaths would be a certainty. It would be like a single engine commercial airliner where gliding was for some reason not possible.

Not having all your eggs in one basket is one of the most key safety approaches.

 

[bahnause]

I have always wondered if this was the case particularly on older funiculars. If it indeed is, I will now be much more reluctant to use them; this to me is a serious design flaw and places an unacceptable risk of failure on one component, where if a failure occurred deaths would be a certainty. It would be like a single engine commercial airliner where gliding was for some reason not possible.

Not having all your eggs in one basket is one of the most key safety approaches.

This is particularly true of modern systems such as Cardada, which have a continuously spliced traction rope. These are much easier to check, and they allow the gondolas to be moved along the rope in order to distribute mechanical stress.

All installations with only one rope have no fallback level anyway.

 

[Egg Centric]

I took the Cardada cable car this morning to go for a hike. It was built in 2000, has two 40-seater cabins and no catch brake. So if the cable breaks, the cabin will go straight down into the valley.

Not everything that is technically feasible or practical is implemented; as always, a risk assessment is conducted.

While imo cable cars probably should have redundancy, there are at least significant practical and financial challenges in providing it. I don't see why that should be the case with providing a funicular with a braking system independent of the cable. Lifts manage it!

I'm one of the most laissez faire on this forum about risks on transportation systems so if I find it concerning it probably is. I'd bet good money it's rebuilt with an independent braking capability.

 

[bahnause]

While imo cable cars probably should have redundancy, there are at least significant practical and financial challenges in providing it. I don't see why that should be the case with providing a funicular with a braking system independent of the cable. Lifts manage it!

There are two braking systems independent from the cable. Neither of them brought the vehicles to a standstill. Whether this was intentional or due to defects or other factors is unclear. My interpretation of the report leads me to doubt whether the brake was suitable for stopping the vehicle, given that the lower vehicle too only came to a standstill at the end of the track.

 

[Egg Centric]

There are two braking systems independent from the cable. Neither of them brought the vehicles to a standstill. Whether this was intentional or due to defects or other factors is unclear. My interpretation of the report leads me to doubt whether the brake was suitable for stopping the vehicle, given that the lower vehicle too only came to a standstill at the end of the track.

Me too and that's why I consider it poorly designed. There's a myriad of designs they could use where the brake would stop the vehicle and they're not complicated or expensive either.

 

[neonison]

There are two braking systems independent from the cable. Neither of them brought the vehicles to a standstill. Whether this was intentional or due to defects or other factors is unclear. My interpretation of the report leads me to doubt whether the brake was suitable for stopping the vehicle, given that the lower vehicle too only came to a standstill at the end of the track.

I’d suspect that the lower car had started moving uphill normally, there was no expectation of an imminent failure and little or no time for any brake to halt the sudden downhill movement.

 

[Bikeman78]

I’d suspect that the lower car had started moving uphill normally, there was no expectation of an imminent failure and little or no time for any brake to halt the sudden downhill movement.

If the brakes couldn't even manage to stop the car that was initially moving in the opposite direction, the downhill car had absolutely no chance. It is odd how we tolerate different risks in different settings though. Like other posters, I find this accident quite unsettling, even though it's extremely rare. Yet I don't think twice about whizzing down a similar gradient on my bicycle.

 

[Bletchleyite]

If the brakes couldn't even manage to stop the car that was initially moving in the opposite direction, the downhill car had absolutely no chance. It is odd how we tolerate different risks in different settings though. Like other posters, I find this accident quite unsettling, even though it's extremely rare. Yet I don't think twice about whizzing down a similar gradient on my bicycle.

Don't know about yours, but my bicycle has brakes, and I wouldn't be whizzing down it faster than they would prevent me hitting a wall.

 

[Bikeman78]

Don't know about yours, but my bicycle has brakes, and I wouldn't be whizzing down it faster than they would prevent me hitting a wall.

Of course it has brakes. My current bike has hydraulic brakes, but in the past I had cables which did snap occasionally. Usually when someone stepped or drove in front of me unexpectedly. As far as I can recall, it always happened on level roads so it wasn't a massive problem. It would be a different story on a steep gradient though.

 

[edwin_m]

I’d suspect that the lower car had started moving uphill normally, there was no expectation of an imminent failure and little or no time for any brake to halt the sudden downhill movement.

The two cars had similar mass, the same braking system and were probably on the same gradient. It is suggested the brakes functioned as intended. So if the cable parted neither car would have had enough brake force to prevent it running away, although the lower one would also be retarded by the friction and inertia of the cable it was still attached to. I'm thinking this could have been even worse if the cars had been at the one-third and two-thirds positions - lower car could have smashed through whatever end of line measures were provided at the bottom and out into the busy square, with the upper one not fast enough to derail on the curve but following into the square a few seconds later.

 

[Bovverboy]

A secondary rotor isn't practical.

Exactly.

A brake that's capable of holding / stopping something on an 18% gradient on eg a rack and pinion type system clearly is.

A rack railway is a rack railway, a funicular a funicular. But yes, you could probably have a compromise between the two.

 

[bahnause]

A rack railway is a rack railway, a funicular a funicular. But yes, you could probably have a compromise between the two.

The original system in 1885 used a rack and water ballast system, in which the car roofs were alternately filled with water to move the trains by gravity. In 1886, steam power was introduced, before it was finally electrified in 1915.