Do trains ever slide past red lights?

[Howardh]

Reading the trains not stopping in the right places at stations thread, made me wonder if, despite the driver on seeing a red signal, attempting to stop in good time, but the train not gripping and sliding past, maybe downhill in icy rails, (a) what sanctions - if any - would a driver face (ie can he prove it "slid") and secondly how would he get to see when the light changed, I assume he would call the signallers?

How often, if at all, does this happen?

 

[dk1]

Yes I've done it. Quite a hairy experience. OTMR download proves everything.

 

[Bald Rick]

Reading the trains not stopping in the right places at stations thread, made me wonder if, despite the driver on seeing a red signal, attempting to stop in good time, but the train not gripping and sliding past, maybe downhill in icy rails, (a) what sanctions - if any - would a driver face (ie can he prove it "slid") and secondly how would he get to see when the light changed, I assume he would call the signallers?

How often, if at all, does this happen?

Yes it happens, particualrly in autumn. The Salisbury incident being one major example

 

[43096]

Reading the trains not stopping in the right places at stations thread, made me wonder if, despite the driver on seeing a red signal, attempting to stop in good time, but the train not gripping and sliding past, maybe downhill in icy rails, (a) what sanctions - if any - would a driver face (ie can he prove it "slid") and secondly how would he get to see when the light changed, I assume he would call the signallers?

How often, if at all, does this happen?

See the Salisbury Tunnel accident report for an example: https://www.gov.uk/government/news/...passenger-trains-at-salisbury-tunnel-junction

 

[hexagon789]

Look up One Bravo Four Seven on YouTube, a slightly dated but relevant training video on 'sliding' past a red.

 

[swt_passenger]

Esher and Hampton Court Junction was another long slide up the mainline in low adhesion conditions, 1A12 passed two signals at danger back in 2005. It’s worth reading the RAIB report. Signaller and driver were both commended for their actions:

https://assets.publishing.service.gov.uk/media/547c906640f0b602410001b3/R252006_070108_Part_1_Esher.pdf

During this incident the train coming off the Hampton Court branch would have likely been hit if the signaller had not rerouted it.

 

[Howardh]

Thanks for the informative answers!

 

[QueensCurve]

Reading the trains not stopping in the right places at stations thread, made me wonder if, despite the driver on seeing a red signal, attempting to stop in good time, but the train not gripping and sliding past, maybe downhill in icy rails, (a) what sanctions - if any - would a driver face (ie can he prove it "slid") and secondly how would he get to see when the light changed, I assume he would call the signallers?

How often, if at all, does this happen?

Yes.

 

[Jamesrob637]

I wasn't aware that Salisbury could've been more catastrophic had it occurred literally 60-90 seconds earlier.

 

[dk1]

Thanks for the informative answers!

In any situation drivers will ALWAYS contact the signaller immediately.

 

[Eoch]

Esher and Hampton Court Junction was another long slide up the mainline in low adhesion conditions, 1A12 passed two signals at danger back in 2005. It’s worth reading the RAIB report. Signaller and driver were both commended for their actions:

https://assets.publishing.service.gov.uk/media/547c906640f0b602410001b3/R252006_070108_Part_1_Esher.pdf

During this incident the train coming off the Hampton Court branch would have likely been hit if the signaller had not rerouted it.

Not quite correct. During this incident, the conficting train from Woking would have likely been hit if the signaller had stopped the train. The train from Hampton Court could not be rerouted at this junction.

 

[swt_passenger]

Not quite correct. During this incident, the conficting train from Woking would have likely been hit if the signaller had stopped the train. The train from Hampton Court could not be rerouted at this junction.

Yes you’re right, I’ve amended the post. Nevertheless it was a clever decision by the signaller?

 

[Y Ddraig Coch]

Yes you’re right. Nevertheless it was a clever decision by the signaller?

and the fact he managed to re set the points also so when 1A12 slid past it didn't cause any damage to the points, the signaler was on top form that day to do so much in such a short space of time under pressure.

 

[norbitonflyer]

and secondly how would he get to see when the light changed, I assume he would call the signallers?

If he has gone past the signal, it will stay at red anyway, because the track circuits will detect his train and thus identify the section protected by the signal as occupied. In the first instance he should contact the signallers who will tell him when it is safe to proceed. Nowadays they have radio, but in the past the driver would have to walk to the nearest signal with a telephone (usually the one he has passed, unless he has gone a long way past it)

In some situations there may be permission to "drive on sight" to the next signal and report there - that means driving at a speed slow enough to stop short of any obstruction (like another train) - but this is only likely to be allowed on plain line - points can be unlocked and therefore able to be moved when a section is protected by a red signal.

 

[Howardh]

If he has gone past the signal, it will stay at red anyway, because the track circuits will detect his train and thus identify the section protected by the signal as occupied. In the first instance he should contact the signallers who will tell him when it is safe to proceed. Nowadays they have radio, but in the past the driver would have to walk to the nearest signal with a telephone (usually the one he has passed, unless he has gone a long way past it)

In some situations there may be permission to "drive on sight" to the next signal and report there - that means driving at a speed slow enough to stop short of any obstruction (like another train) - but this is only likely to be allowed on plain line - points can be unlocked and therefore able to be moved when a section is protected by a red signal.

Is there any system anywhere where when a signal goes red the train is automatically stopped regardless of driver input before that signal?

 

[chuff chuff]

Look up One Bravo Four Seven on YouTube, a slightly dated but relevant training video on 'sliding' past a red.

Ah yeah remember that one very well.

 

[Viper]

Is there any system anywhere where when a signal goes red the train is automatically stopped regardless of driver input before that signal?

TPWS (Train Protection Warning System) does this. Automatically applies the brakes to the train if it is approaching a signal at danger too fast, or if it has passed a signal at danger. I'm not sure how many signals have this in place though, I suspect only signals that protect areas of particular danger?

 

[TurboMan]

TPWS (Train Protection Warning System) does this. Automatically applies the brakes to the train if it is approaching a signal at danger too fast, or if it has passed a signal at danger. I'm not sure how many signals have this in place though, I suspect only signals that protect areas of particular danger?

TPWS won't apply the brake immediately the signal reverts to danger (which is what I think the question was referring to), only when the train passes over the TPWS train stop sensor at the signal (if fitted), or at the overspeed sensor (if fitted, and the train is travelling faster than the trip speed).

In ETCS Level 2, if the signaller withdraws the Movement Authority in an emergency, then the train will enter Trip Mode and the brake will apply immediately.

 

[The Puddock]

In some situations there may be permission to "drive on sight" to the next signal and report there - that means driving at a speed slow enough to stop short of any obstruction (like another train) - but this is only likely to be allowed on plain line - points can be unlocked and therefore able to be moved when a section is protected by a red signal.

No, that's not how the signalling system works.

 

[Zamracene749]

Perhaps not past a red, but a particularly nasty slide happened at Rudine, Croatia in 2009, caused by the application of a new brand of fire retardant to the tracks on a steep downhill section. Train accelerated despite full braking then derailed at high speed. The following rescue train also lost control and derailed, narrowly missing emergency workers.

 

[bahnause]

In ETCS Level 2, if the signaller withdraws the Movement Authority in an emergency, then the train will enter Trip Mode and the brake will apply immediately.

...if there is a GSM-R connection established.

 

[3141]

I wasn't aware that Salisbury could've been more catastrophic had it occurred literally 60-90 seconds earlier.

Or, had it occurred 60 seconds later, there would likely have been no collision at all.

When these incidents happen, they may be absolutely awful, or they may luckily have no serious consequences. The outcome may be the result of human action, good or bad, or or of circumstances outside human control at the time (though, as a result, they may be modified later in an attempt to stop the same thing happening again). It's a reminder that our control of the world is not as complete as we might like it to be.

 

[TurboMan]

...if there is a GSM-R connection established.

Well yes that's true! But even if there isn't, on the Thameslink and GWML ETCS implementations, T_NVCONTACT = 30, and M_NVCONTACT = service brake, so if the RBC connection is lost then the service brake will apply after 30 seconds. One would hope that this would bring the train to a stand in the unlikely event that the signaller needs to withdraw the MA from a train at the same time as the RBC connection is lost.

 

[bahnause]

Well yes that's true! But even if there isn't, on the Thameslink and GWML ETCS implementations, T_NVCONTACT = 30, and M_NVCONTACT = service brake, so if the RBC connection is lost then the service brake will apply after 30 seconds. One would hope that this would bring the train to a stand in the unlikely event that the signaller needs to withdraw the MA from a train at the same time as the RBC connection is lost.

We have a generous 40 seconds on the high speed line. It is usually enough time to get you into the next GSM-R cell (if you are going fast enough).

 

[snowball]

Perhaps not past a red, but a particularly nasty slide happened at Rudine, Croatia in 2009, caused by the application of a new brand of fire retardant to the tracks

Presumably intended for the sleepers?

Is application of fire retardant to tracks or sleepers commonplace?

 

[whoosh]

I remember reading the Control Log in about 2001, where a train slid past a signal at danger due to 'low adhesion/leaves on the line'. The black box confirmed the driver's report that he'd approached very slowly... - the train had gone over the AWS magnet on approach at 5mph, and with the brake still in, and with an emergency application, had still SPADded.

I think the terminology/classification system has been changed now, but a there are categories for:

• Driver's fault (I.e. a misjudgement of braking)
• Driver behaved as expected, but train didn't (I.e. a slide in very low adhesion - that's adhesion that is lower than expected at that location in the weather conditions on the day, or a fault with the train)
• Signal is replaced to danger in front of driver, so can't stop in time. This can be a Signaller's error, a purposeful act by the Signaller to prevent an incident, or a failure of equipment. I've had several of these over the years, including flooding and a lightning strike reverting signals to danger by the signals 'failing safe' to danger when affected by short circuits.
• Runaway vehicles passing a signal at danger.

These used to be Categories A - D, with B being the slide example, but I can't remember what they are called now.
The old category A was the driver's fault and went against them on their record. A category D could also, if a parking brake on a loco or unit hadn't been set by a driver. B and C weren't.

A black box will show train speed, what braking inputs were made and when, whether the sander was activated, and examination of the line will show if sand was actually deposited. Modern trains will show a fault on the computer system if the sand boxes are low on sand.

Generally, management of railhead treatment is much better these days than twenty years ago in the story I mentioned above, although occasionally an incident will still occur, such as the Salisbury one.

 

[43066]

In some situations there may be permission to "drive on sight" to the next signal and report there - that means driving at a speed slow enough to stop short of any obstruction (like another train) - but this is only likely to be allowed on plain line - points can be unlocked and therefore able to be moved when a section is protected by a red signal.

No, that's not how the signalling system works.

“Proceed at caution to signal no. X at a speed at which you’ll be able to stop short of any obstruction” etc., with the entire message to be read back by the driver to ensure no misunderstanding, would be a perfectly valid instruction. Whether it would necessarily be required following a SPAD is another question - potentially if the track circuits were playing up, or in absolute block non track circuited areas it might, but only if the driver felt able to continue.

 

[The Puddock]

“Proceed at caution to signal no. X at a speed at which you’ll be able to stop short of any obstruction” etc., with the entire message to be read back by the driver to ensure no misunderstanding, would be a perfectly valid instruction. Whether it would necessarily be required following a SPAD is another question - potentially if the track circuits were playing up, or in absolute block non track circuited areas it might, but only if the driver felt able to continue.

It was the bit about points being 'unlocked' I was taking issue with.

 

[Bald Rick]

I remember reading the Control Log in about 2001, where a train slid past a signal at danger due to 'low adhesion/leaves on the line'. The black box confirmed the driver's report that he'd approached very slowly... - the train had gone over the AWS magnet on approach at 5mph, and with the brake still in, and with an emergency application, had still SPADded.

I think the terminology/classification system has been changed now, but a there are categories for:

• Driver's fault (I.e. a misjudgement of braking)
• Driver behaved as expected, but train didn't (I.e. a slide in very low adhesion - that's adhesion that is lower than expected at that location in the weather conditions on the day, or a fault with the train)
• Signal is replaced to danger in front of driver, so can't stop in time. This can be a Signaller's error, a purposeful act by the Signaller to prevent an incident, or a failure of equipment. I've had several of these over the years, including flooding and a lightning strike reverting signals to danger by the signals 'failing safe' to danger when affected by short circuits.
• Runaway vehicles passing a signal at danger.

These used to be Categories A - D, with B being the slide example, but I can't remember what they are called now.
The old category A was the driver's fault and went against them on their record. A category D could also, if a parking brake on a loco or unit hadn't been set by a driver. B and C weren't.

A black box will show train speed, what braking inputs were made and when, whether the sander was activated, and examination of the line will show if sand was actually deposited. Modern trains will show a fault on the computer system if the sand boxes are low on sand.

Generally, management of railhead treatment is much better these days than twenty years ago in the story I mentioned above, although occasionally an incident will still occur, such as the Salisbury one.

The categories changed a few years ago.

Essentially, SPADs are where a driver goes past a signal that was intentionally red and red in sufficient time for the driver to stop.

Where a signal is red unintentionally, or where a signal is ’put back’ in front of a driver such that he/she cannot stop, then that is a SPAR (Signal passed at red).

SPADs are classified in 5 categories, which I’ll detail another time.

 

[43066]

It was the bit about points being 'unlocked' I was taking issue with.

Understood.

You would still (IIRC) be instructed to proceed at 15mph max over junctions, level crossings etc. (I should know for certain, but can’t remember off the top of my head).

Where a signal is red unintentionally, or where a signal is ’put back’ in front of a driver such that he/she cannot stop, then that is a SPAR (Signal passed at red).

Formerly known as a Category 2 SPAD, I do believe.