Level crossing barriers going down - questions (specifically Hest Bank)

[newWCMLwatcher]

Hi all,

Many apologies if I'm asking a duplicate question or if I'm posting in the wrong place. I have for many years wondered what triggers the barriers to go down on CCTV controlled level crossings. My local one is Hest Bank on the WCML.

I've often wondered various things such as the below:

• Is there a specific time limit before the train passes that the barriers must be lowered? If this guidelines or rule?
• At what point in the track do the barriers lower if auto? If auto, what's the trigger - speed, mileage, etc?
• Does Preston PSB lower at their manpower, or is it automated? Warning bell and manual action or do they go down on their own?
• Is Hest Bank and Bolton-le-Sands co-ordinated to operate simultaneously? i.e. same time
• When a train passes, what constitutes a clear passing and therfore raise of the barriers?
• If a train reversed out of a qualifying section (if such exists) would the barriers lift?

Thanks so much, and apologies for so many questions (daft ones at that I'm sure you're thinking).

Many thanks

 

[newWCMLwatcher]

Also - sorry! Does the "rule" change depending on type of train or is it a blanket approach for all?

 

[civ-eng-jim]

Someone definitely will.....but you may have to wait for a response when it's not in the middle of the night.

 

[jopsuk]

as far as I understand the timing with CCTV crossings is such that (as long as there's no other factors) the driver will see only green signals on approach and hence can drive as fast as is permitted.

 

[Ianno87]

The signaller in control of the crossing basically just aims for an approaching train never to see anything other than green signals. I suspect the training for each crossing includes the recommended point at which to start the process, though I'd expect personal taste to dictate how 'agressive' individual signallers will be. A few crossings have 'alerts' on approach, i.e. a tone will sound when a particular track circuit is occupied, to get the signaller's attention.

Observing Chesterton Level crossing (just south of Cambridge North), the sequence typically starts for a Down/Northbound train when it has departed and left the platform at Cambridge, for example (the platform starters being a Red-Yellow-Green distance away)

 

[Ducatist4]

I'd assumed the whole process was automated and the signaller just observes the CCTv to make sure the crossing is clear?

 

[Tomnick]

Hi all,

Many apologies if I'm asking a duplicate question or if I'm posting in the wrong place. I have for many years wondered what triggers the barriers to go down on CCTV controlled level crossings. My local one is Hest Bank on the WCML.

I've often wondered various things such as the below:

• Is there a specific time limit before the train passes that the barriers must be lowered? If this guidelines or rule?
• At what point in the track do the barriers lower if auto? If auto, what's the trigger - speed, mileage, etc?
• Does Preston PSB lower at their manpower, or is it automated? Warning bell and manual action or do they go down on their own?
• Is Hest Bank and Bolton-le-Sands co-ordinated to operate simultaneously? i.e. same time
• When a train passes, what constitutes a clear passing and therfore raise of the barriers?
• If a train reversed out of a qualifying section (if such exists) would the barriers lift?

Thanks so much, and apologies for so many questions (daft ones at that I'm sure you're thinking).

Many thanks

More or less in order:

There isn't a specific minimum time between barriers down and train passing. The crossing is protected by signals at danger until the barriers are down and the crossing is observed to be clear (there's a 'crossing clear' button on the crossing controls, which confirms that the signalman or crossing keeper has checked it). Normally, though, you don't want the train to have to slow down, so you need the barriers down in good time before the driver reaches the first restrictive aspects (two signals back from the protecting signal on a four-aspect line, so for Hest Bank crossing, you're probably looking at getting the signals cleared before a Down fast train passes Lancaster and on the Up before it passes Carnforth).

Some crossings do have an auto-lower facility which (if turned on and if there's a route set over the crossing) will start the lowering sequence when the train occupies a pre-determined track circuit, that track circuit being such that there's time for the barriers to come down and the signalman to observe the monitors in plenty of time to press the 'crossing clear' button and clear the signals to achieve the above.

These crossings don't appear to have auto-lower though, so it's all done at the signalman's discretion. There's a small panel with the controls for these two crossings and Bare Lane with its own dedicated operator (not sure whether they'll be classed be a signalman or a crossing keeper?). Incidentally, the routes are set by the signalman on the main panel, this panel controls 'slots' on the protecting signals, meaning that they won't clear until the main panel man has a route in *and* the crossing panel man has done the business and cleared the slot.

They don't operate simultaneously - they each have their own controls and will be operated as necessary to keep both trains and road traffic moving as efficiently as possible.

As soon as the train passes clear of the track circuit covering the crossing (usually just a few yards beyond it), the barriers will be free to raise provided that there's not a train signalled over on the opposite line. Most CCTV crossings have an auto-raise facility, so (again, if turned on - there are times when you don't want the barriers to come up of their own accord!) it'll do this itself.

I don't know, to be honest, how auto-raise copes with a train setting back *after* passing the protecting signal nor what happens if the protecting signal is replaced to danger (with the slot replaced) with a train on the approach - perhaps because the train has failed. I'd imagine that they'd have the auto-raise turned off by that stage! Generally, though, you need the protecting signal (or slot) at danger and all track circuits between the signal and the crossing clear to allow the barriers to raise. If the protecting signal is replaced to danger with a train approaching, there'll be a pre-determined time delay (usually something like two minutes, but it depends on various factors and will be designed to ensure that an approaching train either has time to stop short of the signal *or* pass it at danger and thus hold the crossing down by occupying track circuits in the route instead) before the route releases and the barriers can be raised.

There's some excellent photos of Preston PSB here, including detail views of Hest Bank and its protecting signals on the main panel, an overview of the crossing panel and a detailed view of the controls for Hest Bank crossing.

 

[newWCMLwatcher]

Many thanks to everyone for the replies. Thanks @Tomnick for providing all that information, it's really interesting to me. I guess it will be second nature to experienced signallers with their timings and it certainly sounds like it needs some skill to time it well enough so as not to lower the barriers too early but to keep the train running at line speed.

 

[MarkyT]

I don't know, to be honest, how auto-raise copes with a train setting back *after* passing the protecting signal nor what happens if the protecting signal is replaced to danger (with the slot replaced) with a train on the approach - perhaps because the train has failed. I'd imagine that they'd have the auto-raise turned off by that stage! Generally, though, you need the protecting signal (or slot) at danger and all track circuits between the signal and the crossing clear to allow the barriers to raise. If the protecting signal is replaced to danger with a train approaching, there'll be a pre-determined time delay (usually something like two minutes, but it depends on various factors and will be designed to ensure that an approaching train either has time to stop short of the signal *or* pass it at danger and thus hold the crossing down by occupying track circuits in the route instead) before the route releases and the barriers can be raised.

Trains passing over the crossing must clear the track circuit over the road and operate treadles in the correct order and timing to trigger auto raise, so if a train stops over the road and sets back, the barriers should remain locked down until the road is clear. Approach locking can be 'when operated' which means the timeout occurs whenever the signal is replaced, or 'comprehensive' where the route ahead is unlocked immediately if there is no train approaching when the signal is replaced. Having a crossing in the section is usually grounds to provide comprehensive approach locking at a signal even in an interlocking where 'when operated' applies elsewhere. 'Comprehensive' allows the road to be reopened to road traffic more quickly after a train fails to depart after being signalled.

 

[Belperpete]

Just to add to what Tomnick said:

Once the crossing close has been triggered (whether by the signaller or automatically), the sequence of lights, 'bells' and barriers is entirely automatic, although the signaller can stop the sequence, e.g. if he spots someone trapped on the crossing.

Where auto-lower is provided, it takes no account of the type or speed of the approaching train. As Tomnick says, it is designed so that the crossing can close, the signaller press the "crossing clear" button, and the sequence of signals on the approach to the crossing all clear up to green, before the train gets to the sighting point of the first 'distant' signal. A fixed "strike-in point" is defined: auto-lower will be triggered when a train is detected passing this point (provided auto-lower is switched on and the route is set over the crossing), regardless of the speed or type of train.

Having said that, where there is a station on the approach to a crossing, then a "stopping/non-stopping" control may be provided. Normally, the crossing closure (and clearing of the protecting signal) will be delayed to allow for the train stopping at the station, which means that the train will get cautionary aspects. However, if the signaller presses the non-stopping button, then the controls will close the crossing earlier, so that the train does not get cautionary aspects. In such cases, there will be stopping and non-stopping strike-in points.

Auto-raise requires that a train has been signalled over the crossing. If the crossing has been closed, and the signaller wants to re-open it before a train has passed over, then the signaller will need to operate the manual raise control to re-open the crossing. For example, if a train fails on the approach to the crossing, and the signaller cancels the route (which will require an approach locking "time-out" of several minutes, as explained by MarkyT), the signaller will then need to manually open the crossing.

If an approaching train is going to reverse back away before passing over the crossing, then the signaller should not have set the route for the train to pass over the crossing, and so the crossing should not have been closed in the first place. However, if for some reason he had set the route and the crossing had closed, then as before the signaller would need to manually cancel the route (which will likely require the approach locking to time-out) and then manually re-open the crossing.

How fast auto-raise happens depends on how close the track-circuit joints are to the edge of the crossing - the further away they are, and the longer the back of the train takes to pass over them, then the longer it will take for the crossing to re-open after the train.

Auto-raise may not work for all routes. For example, it is not usually provided for shunt moves. So if you have a trailing crossover, and a train sets back over this crossover and over the crossing on a shunt route, then the signaller may have to manually re-open the crossing.

Again, once crossing re-opening has been triggered, the sequence is entirely automatic - all four barriers will rise together, but the red lights will keep flashing until the barriers have raised. Once the barriers have started rising, then auto-lower is inhibited until the road has been open for a minimum time (usually 10 seconds).

If an unsignalled train is detected approaching the crossing, then the crossing will go into an emergency-close mode. This may go straight to red flashing lights without the steady yellow lights = it won't, however, lower the barriers. So, for example, if a train over-runs the protecting signal at red, the crossing will go into this emergency-close mode. Where the protecting signal is very close to the crossing, additional over-run treadles may be provided specifically to detect a train over-running toward the crossing. Auto-raise won't work if the over-running train passes right over the crossing, as it wasn't signalled over - the signaller will have to manually re-open the crossing.

 

[edwin_m]

Approach locking can be 'when operated' which means the timeout occurs whenever the signal is replaced, or 'comprehensive' where the route ahead is unlocked immediately if there is no train approaching when the signal is replaced. Having a crossing in the section is usually grounds to provide comprehensive approach locking at a signal even in an interlocking where 'when operated' applies elsewhere. 'Comprehensive' allows the road to be reopened to road traffic more quickly after a train fails to depart after being signalled.

However, comprehensive approach locking only releases the route immediately if no approaching train could have seen a signal that will change aspect when that happens. So if the route is set over the crossing for an approaching train to run on greens and then needs to be cancelled, it's quite likely that the train will by than be close enough that it will revert to the timer.

Lack of approach locking at a level crossing was the cause of a serious accident at Moreton-on-Lugg in 2010, when the signalman got confused about where the train was, thought it had passed, replaced the signals and opened the crossing to a road vehicle which was then hit by the train. Approach locking was fitted to most such crossings but not this one, and would have allowed the signal to be replaced but prevented the opening of the crossing for a timeout period.

 

[Llanigraham]

However, comprehensive approach locking only releases the route immediately if no approaching train could have seen a signal that will change aspect when that happens. So if the route is set over the crossing for an approaching train to run on greens and then needs to be cancelled, it's quite likely that the train will by than be close enough that it will revert to the timer.

Lack of approach locking at a level crossing was the cause of a serious accident at Moreton-on-Lugg in 2010, when the signalman got confused about where the train was, thought it had passed, replaced the signals and opened the crossing to a road vehicle which was then hit by the train. Approach locking was fitted to most such crossings but not this one, and would have allowed the signal to be replaced but prevented the opening of the crossing for a timeout period.

And following that incident many Boxes were modified to add the equipment to stop this happening.
Mine was one of them.

 

[newWCMLwatcher]

Thank you all for sharing more detail on this topic. I do find it fascinating and it's good to know more about it. I knew that somehow the crossing would need to be linked to the signalling system but I naively thought it was a much simpler concept that it is! I had always thought for example, that the barrier close process was triggered automatically when a train passes a certain point in the line, and a signaller is notified by an alarm that the closing is crossing and to watch the CCTV - if everything was okay, then no action required. Then once the signaller saw the train pass on CCTV, or saw the following signal section occupied, would open the crossing. Thinking about it more, I knew that would be too simplistic so that's why I asked.

I've always looked at crossings for sensors and equipment to see if I could work out what would confirm a train has passed, but I now understand it's all very cleverly done with circuits and as technology has improved over the years, I bet you can't even see most of the wiring and equipment needed to make this all happen.

Over the last few years my interest in the infrastructure has increased so it's good to be able to understand more about it - thank you to everyone for sharing their knowledge on the subject.

 

[edwin_m]

I've always looked at crossings for sensors and equipment to see if I could work out what would confirm a train has passed, but I now understand it's all very cleverly done with circuits and as technology has improved over the years, I bet you can't even see most of the wiring and equipment needed to make this all happen.

At most level crossings unless re-signaled in the last decade or so, detection will be primarily by track circuit. These rely on the wheels of a train making an electrical connection between the two rails, so there is no sensor as such. You may see insulated joints in the rails and cables connecting them to trackside cabinets. However at level crossings they are usually also supplemented by a treadle, which is a small box close to the inner side of the rail with a rod sticking out which is pushed down by the wheels of a passing train. This is in case the track circuit is momentarily defeated by contaminated rails, which might delay the start of the light sequence.

More modern crossings may have detection via axle counters, which have what is sometimes described as a double mushroom on the inside of the rail. There is a metal detector in each mushroom and each passing wheel activates one and then the other in quick succession. The system detects this and knows how many wheels have passed and in which direction. By combining the readings from the detectors at each end of a track section it can work out whether all the wheels that have gone in at one end have come out at the other.

 

[Belperpete]

I had always thought for example, that the barrier close process was triggered automatically when a train passes a certain point in the line, and a signaller is notified by an alarm that the closing is crossing and to watch the CCTV - if everything was okay, then no action required. Then once the signaller saw the train pass on CCTV, or saw the following signal section occupied, would open the crossing. Thinking about it more, I knew that would be too simplistic so that's why I asked.

You were not too far off the mark. The barrier close process is triggered automatically when a train passes a certain point in the line, and the signaller is notified by an alarm that the closing is crossing and to watch the CCTV (in some places the CCTV turning on is sufficient to draw the signallers' attention). The only thing you really got wrong was the "no action" bit - the signaller is required to confirm "crossing clear" before the signal(s) will clear. Again, your bit about re-opening the crossing is not far off the mark for the many crossings that don't have auto-raise: when the signaller sees that the train has passed (either from his signalling panel or from the CCTV), he will re-open the crossing.

The CCTV isn't usually left turned on, as that could be distracting for the signaller - it is normally turned on when the crossing closure process starts. For crossings with manual raise, the CCTV automatically turns off when the crossing has re-opened, for those with auto-raise, the CCTV turns off as soon as the signaller presses crossing-clear, as the raising process is entirely automatic.

I should point out that all the above applies for full-barrier crossings monitored by CCTV. Full-barrier crossings with automatic obstacle detection are very different. And the two types of automatic crossings with half-barriers are very different too.