Accident at Alton Towers

[Mysterious Sue]

I thought it was worth summarising again what we know:

- they were having technical problems so ran at least one test train, which valleyed.
- then they ran a train full of people which reportedly stopped at the top of lift hill (as it should do under normal safety procedures) for 10 to 15 minutes. It did not stop on the inversion, this is impossible and only reported by the Daily Fail

...and what we are debating:

- Something triggers the second train to start again and allow two trains into the same area and crash.



The way I see it, there are 3 possible reasons (this is all from my head and I'm terribly uninformed, but I'm just trying to make sense of it):

A. The ride-ops 'allowed this' by sending a train when they weren't meant to/weren't told of the valleyed train by the system. I don't imagine this is possible but I don't know enough to rule it out.
B. Engineers reset the system and 'cleared the ride of errors' so that the valleyed train 'disappeared' and the second train was free to start again. For a 10 to 15 minute breakdown, surely engineers would start to be on the scene?
C. The ride started again of its own accord with no human input. If this is the case then Gerstlauer will be at fault and we may see other rides being shut down to have their systems checked.

 

[Ian]

^ Good point, do we know if the other Infinity Coasters have been closed for checks? Or are we treating this as a lone Smilier incident?

 

[Nemesis Inferno]

Dunno about Karacho, but Saw is closed... Whether that's because it's related to this is unknown...

 

[Pink Cadillac]

I think staff were aware of the valleyed train. 10-15 mins is a very long time and when I was there two weeks ago on the Monday, there were a lot of other staff all over x-sector. one of them must have spotted it.

 

[Smithy]

Aye but the ride-ops inside the building may not necessarily have been aware.

I have no idea how many trains were on track but wouldn't they have had to clear the train in the station to clear a block space for the train that ultimately valleyed to clear the second lift block? I know it's mostly media ignorance and hysteria but I don't think they sent the car out while the train was already valleyed

 

[Slayed]

If I ran a park with a Gerstlauer coaster, I'd have been on the phone/email yesterday afternoon asking them if there's anything to worry about.

If it was an Infinity model I'd be dispatching one train until further notice! But then the coaster ops I saw at Tripsdrill a while back weren't exactly exhaustive in their safety checks (were they Richard ).

As for the Daily Fail article, I despise the rag - I wouldn't even wipe my arse with it after the world's hottest vindaloo - but that diagram is actually a very good effort to accurately portray what they've been told, even when to those of us with coaster knowledge the idea of a train being "held" inverted is nonsense.

 

[Howie]

I don't have the link, but a hospital spokesperson has said that one of the 4 'seriously' injured people have been discharged from the hospital.
So that's good news then...
However, most of the news articles I've read are still stating that at least one person has sustained 'potentially life changing' injuries. Which does kinda imply limb-loss or some sort of paralysis, so it looks like not everyone has escaped so lightly. Poor bloke. Just horrible.

 

[CoasterCrazy]

^Excellent news for those discharged..but not so much for those who remain in hospital.

OK, so I've had a little think about this. The question I ask is:
What if there were two faults which occurred at almost exactly the same time, one of which was the valley?

I'm not too familiar with control panels, especially the Smiler's which admittedly, I've never seen, but how do the operators know when there's a fault in the system....I'm aware of the 'Trouble' button, but if there were two faults, would they necessarily be aware of both at a glance?
This is my proposed mechanism:

1 - The ops send out the empty train, and in their rush to reopen the ride after its 3 hour closure, dispatch a loaded train very shortly after, not waiting for the empty train to return...possibly a bit of an oversight, but understandable.

2- As we know, for some reason, possibly busting a wheel, or some other unlikely anomaly (considering the ride was already warmed up from testing earlier), and valleys on the batwing.

3- At the moment one train valleys, the loaded train is already on the lift hill. The PLC control system doesn't detect anything, as it's not expecting the empty train at the block yet.

4- There's some kind of problem on the lift hill, possibly a sensor or something minor, and the control system immediately stops the ride. Of course, before, after or at exactly the same time, the PLC detects that the empty train has not made it to the next block, which would have had a similar effect...

5- Detecting the loaded train is stuck on the lift hill, the ride ops assume there to be a problem exclusively on the lift hill, and set about somehow fixing the problem, or not, but in the process resetting the system. What they should do at that moment is evacuate the lift hill, but having had the earlier, lengthy breakdown, the ops are under pressure to reopen the ride, they neglect to do this. Frustrated, they also forget about the empty train they sent out earlier.

6- The reset seems to do the trick - the lift hill restarts and all is well, but the control system no longer notes the ride has valleyed.
Because of the 'vanishing effect' of the empty train on the block system, neither the lift hill, nor the airtime hill trims stop the ride.
As far as the ride ops are concerned, they've just done a good job and prevented the ride closing again.

7- By this time, the GP and staff have noticed that the ride has valleyed - but somehow, they do not communicate the message at all, thinking the ride operators are aware of this, or do not get the message across in time. After all, as the ride has stopped on the lift, it looks like an emergency stop because of the earlier valley, and the ride ops are just about to evac the lift hill, bring the crane out, and close the ride.

8- Of course, at this point, the ride accident occurs, but a third train is already ascending the lift hill before the PLC system detects the loaded train has 'valleyed' and again, shuts the ride down. I believe there was another train evacuated wasn't there? Either way, still works.

This requires so many variables to come off, and is of course, just a combination of very unlikely events. But I suppose this is technically possible (or at least I think it is). That or there was a serious, elementary failure in the PLC system, which would be very worrying...

I think staff were aware of the valleyed train. 10-15 mins is a very long time and when I was there two weeks ago on the Monday, there were a lot of other staff all over x-sector. one of them must have spotted it.

Yeah,you would have thought so! In a few reports I read, the staff were apparently made aware of this by the people queuing...but see event 7.



Anyway, when this eventually reopens, we'll be in a good position:
-Short queue times due to the inexorable 'fear'
-Possibly redesigned or refurbished trains
-They'll probably reprofile some of the track, making the ride smoother and seemingly 'safer'
-And of course, just like the aftermath of the NTG aftermath, the park will OCD on safety, and the amendments they make, and in that first period, with extra precaution, the ride will extremely, extremely, safe - and the chance of an accident like this will be 0.


Just remember, they'll need to check for signs of foul play...which for something like this, is actually possible, but unlikely.

 

[Pokemaniac]

If a ride is restarted, is it possible that the lift hill will automatically start up and the train sent by accident? I would really hope not, but a misunderstanding by the engineers on how the ride will respond to a reboot is always an option.

Another theory (throwing them out here now!) could be that Smiler is unique in that it doesn't have anti-rollbacks meaning the trains can be rolled backwards down the lift hill in the event of an evacuation. We've seen this happen in the past and may well be something they were trying to do with people stuck at the top for 15 mins and the knowledge that a car had stalled so an evacuation was inevitable. You then ask the question of if its possible to manually reverse the train, can you also send it forward over the top of the lift by accident? I doubt anyone here will know unless you're a Smiler engineer! :wink:

It seems that the lift hill does not have its own block brake. The "brake" on the lift hill block is simply the chain stopping, and gravity holding the train in place. As such, if the chain were to start, any train resting near the top would be pushed forward and left to gravity's devices - at which point there is nothing to slow or stop the train until it hits the next set of brakes. Judging from POVs, there is no secondary system that could catch and stop the train after the crest of the lift (well, theoretically the trim brakes could interfere, but that would require them being calibrated for that, which is not very practical for normal operations). If the chain is in motion, and the train passes the crest, it will continue 'till it hits the next set of brakes before the second lift.

It could be that the lift would start automatically at some point in the boot procedure, but I think that's unlikely - or a serious design oversight. Testing the chain drive would be a vital part of the morning startup, but unlikely to be automatic, due to the lack of a redundant system to prevent eventual trains on the lift to be rushed out to the track. Then again, this normally shouldn't be problem. After all, the roller coaster is designed to send trains safely around the track. It's only in the exceedingly rare case of an obstruction that sending a train off the lifthill would be dangerous. And the routines everything relies on, explicitly tell the operators to check if everything is ready before the system starts. If this step is neglected, the system would have no way to know if the situation is safe or not. If a step of the procedure is manual and reliant on humans, the system will always assume that everything is in order. It cannot tell otherwise.

Overall, I suspect neglect to be at work. Or a situation where the engineers try to interfere with the lift hill, and as Peter suggested, it goes forwards instead of backwards.

 

[Howie]

Either way, those 16 people stuck on the lift hill were very probaby aware of the stalled train ahead of them - being stuck there for 10 minutes they would have been looking round to see what was going on, taking selfies etc..

"Here's me stuck on Smiler's lift hill - lol".

Can you imagine the sheer terror when the lift hill started going forwards again? And then having to endure another 4 inversions before crunch time?

Jesus. Gives me the proper chills.

 

[mcg0ughchris]

Either way, those 16 people stuck on the lift hill were very probaby aware of the stalled train ahead of them - being stuck there for 10 minutes they would have been looking round to see what was going on, taking selfies etc..

"Here's me stuck on Smiler's lift hill - lol".

Can you imagine the sheer terror when the lift hill started going forwards again? And then having to endure another 4 inversions before crunch time?

Jesus. Gives me the proper chills.

That's exactly what I was thinking.

Alton going to have to shell out some huge well deserved settlements to these 16 people.

The accident can happen but the time frames don't make any sense at all in my opinion. A 10 to 15 minute delay in which the tech staff must have been working on getting the ride up and running again before evacuating guests, all this when knowingly a car had valleyed a short way down the track. Unacceptable protocol.

 

[Gazza]

As far as I know, Infinity coaster vertical lifts are designed so a train can be lowered backwards when an evac is needed, with mag brakes controlling the descent.
Did someone stuff up the evac procedure and advance the train forwards rather than reverse it. Wrong button?

 

[mcg0ughchris]

As far as I know, Infinity coaster vertical lifts are designed so a train can be lowered backwards when an evac is needed, with mag brakes controlling the descent.
Did someone stuff up the evac procedure and advance the train forwards rather than reverse it. Wrong button?

For the train to stop on the lift hill then surely the sensors were working which ruled out a system fault?

Either the system got restarted with the staff thinking that other car had not valleyed and was going to move also.

The staff forgot about the test car or there was a huge error in terms of the evacuating process like you mentioned.

I fail to see how the system could just reset itself and start moving the ride car again? Surely that can't happen as that'd be a huge flaw.

So the responsibility has to fall on human error and the also the fact there wasn't a failsafe in place to detect cars that had valleyed

 

[Smithy]

It seems plausible, there's a few videos around of rollbacks on the vertical lift but it's impossible to see if it rolls forward even slightly (which could be a key point as the pictures of the train on the lift hill show it nearly cresting the edge), in one of them a bloke over the tannoy tells the rollback train and a train waiting directly behind them in the block brakes they've got to reset the lift and as soon as they do that it starts straight back up again.

So many different aspects of this investigation tbh, I think there's a number of factors that have just unfortunately all happened at once to cause this horrible accident. Ride op error in reacting to a computer issue, all of which wouldn't have caused a collision if the ride had been corrected itself when the issue of valleying presented itself last season.

 

[mcg0ughchris]

It seems plausible, there's a few videos around of rollbacks on the vertical lift but it's impossible to see if it rolls forward even slightly (which could be a key point as the pictures of the train on the lift hill show it nearly cresting the edge), in one of them a bloke over the tannoy tells the rollback train and a train waiting directly behind them in the block brakes they've got to reset the lift and as soon as they do that it starts straight back up again.

So many different aspects of this investigation tbh, I think there's a number of factors that have just unfortunately all happened at once to cause this horrible accident. Ride op error in reacting to a computer issue, all of which wouldn't have caused a collision if the ride had been corrected itself when the issue of valleying presented itself last season.

So is it possible that the breaks holding the car on the hill just let the car go when the tech team were trying to sort out the system?

 

[Slayed]

^Seems to be pretty unlikely - if the train was "released" on the lift hill you'd expect it to fall back under gravity not move forwards.

One more interesting speculative comment - I've read a report, from someone in the queue at the time, that prior to the stall it was announced that they were adding another train, so it's possible the stalled train had literally only just been brought onto the track and was trying to complete its first run. Could also explain why the ride may not have been in its usual operating state prior to the incident.

 

[Pink Cadillac]

The lift in question is not the vertical lift and I would presume the train could be evacuated at any time by staff.

I would have thought that this would be one of the first procedures when the coaster becomes out of action, or in this case, when the previous train had valleyed.

 

[mcg0ughchris]

That's what I don't understand pink panther, we've established that the security sensors on the ride recognised that the car had not left the designated block of track and therefore stalled the car carrying people on the lift hill.

If it was a good 10 to 15 mins then surely they must have received word in this time that the car had valleyed and wasn't going to move even if they got the rude up and running. So before touching anything technical they should have released the passengers and went for safety first. This is possible right? Or am I missing something completely?

 

[Jamesss]

It does seem odd that in the entire time of being stuck on the lift for 10/15 mins there was apparently no attempt to evacuate the train. With the previous train stalled in the batwing, there would clearly be no way the ride could continue operating anytime soon so you would assume an evac should've happened right away. I guess this all suggests that the staff didn't even know about the stall occurring.

 

[CoasterCrazy]

^^^Thanks for that...was about to say the same myself! Even if it rolled back off the vertical lift, there's no way it would have been able to get back all the way to the Batwing. I highly doubt they were messing around with the train at the top, especially since the train was fully loaded.

Also, surely in a last resort situation where the computer detects two trains are inevitably going to collide, the trim brakes on the first hill would be activated?

See my post on the previous page!