Level Crossings and interlocking

[ainsworth74]

I was just pondering whilst stood waiting for a level crossing to raise the other night about how level crossings and specifically Manually Controlled Barrier (MCB) crossings (whether they controlled directly by a signaller in the adjacent box or via CCTV or Obstacle Detection) work in terms of interlocking with the signals. My understanding is that for the protecting signal to clear the MCB has to be down and if the signalling system cannot detect the crossing as being down the signals will not clear. Assuming I'm right, so far so good.

My question is more about what happens once the barriers are down and the signal clears. Does that then lock the barrier controls so that they cannot be raised as long as the route is set over the crossing? When a train is in section are barriers locked until the train is clear of the section and the overlap and the route releases? I was just wondering as last night I could see the signaller stood by the controls for the crossing but it took a while for them to actual raise the barriers even after the train had passed which led to me wonder if that was because they had to wait until the train was clear of the overlap and the route had been released?

 

[Llanigraham]

I can only respond in relation to my AB Box, but post- Moreton on Lugg, the barriers could only be raised after the train had passed over the crossing and the treadle had been activated and I had replaced the relevant Home and Section signals back into the frame. Only then could I activate the lift barriers button.

 

[Annetts key]

I was just pondering whilst stood waiting for a level crossing to raise the other night about how level crossings and specifically Manually Controlled Barrier (MCB) crossings (whether they controlled directly by a signaller in the adjacent box or via CCTV or Obstacle Detection) work in terms of interlocking with the signals. My understanding is that for the protecting signal to clear the MCB has to be down and if the signalling system cannot detect the crossing as being down the signals will not clear. Assuming I'm right, so far so good.

My question is more about what happens once the barriers are down and the signal clears. Does that then lock the barrier controls so that they cannot be raised as long as the route is set over the crossing? When a train is in section are barriers locked until the train is clear of the section and the overlap and the route releases? I was just wondering as last night I could see the signaller stood by the controls for the crossing but it took a while for them to actual raise the barriers even after the train had passed which led to me wonder if that was because they had to wait until the train was clear of the overlap and the route had been released?

At a local MCB or CCTV MCB (controlled by a signal box, signalling centre, or by a crossing keeper), before a signal can be cleared, enough of the red road lights have to be lit, all the barriers have to be proved fully lowered, and the operator has to confirm that the crossing is clear by pressing a crossing clear button.

As soon as a protecting signal shows a proceed aspect (or tries to), approach/route locking is activated (*). This now locks the crossing controls. The operator now will be unable to raise the barriers.

The approach/route locking will remain effective until either, the interlocking sees the correct sequence of track circuits (or axle counters) being operated (that shows that the train has passed the protecting signal, in other words, normal passage of the train), the signaller cancels the signal route and the interlocking can see via the track circuits (or ACs) that there is no train in section or approaching (comprehensive approach locking)(if provided), or the route has been cancelled and the protecting signal is now red and then after a time delay (normally 2 or 3 minutes).

After the train passes the protecting signal, and until it is clear of the crossing, track circuits (or ACs) directly lock the crossing controls.

Normally, the barriers can be raised and the crossing opened to road traffic as soon as the train has cleared the track circuit (or AC) that is over the crossing assuming that the switch/lever/control for the signal has been restored to the normal (on/danger) position. You do not have to wait until the train has cleared the complete signal section.

The above is the general principles for unidirectional lines. It’s more complex for single lines, bidirectional lines or if points/junctions are involved.

* Note, actual practice has varied slightly over the years, there are many differences between different level crossing controls. Hence some have separate route locking and approach locking, while others combine it together. The end result is the same, the crossing controls are locked as described above.

 

[Llanigraham]

As soon as a protecting signal shows a proceed aspect (or tries to), approach/route locking is activated (*). This now locks the crossing controls. The operator now will be unable to raise the barriers.

That certainly would be correct now, but prior to the Moreton-on-Lugg incident that didn't apply at my Box, or others near-by. I had to prove to an RAIB inspector what we could do when he visited my Box in the week after that, and it was why my Box and others were fitted with treadles to lock the barriers until a train had passed, or to give us 5 minutes "thinking time" if we did want to raise the barriers for any reason if one hadn't.

 

[ainsworth74]

At a local MCB or CCTV MCB (controlled by a signal box, signalling centre, or by a crossing keeper), before a signal can be cleared, enough of the red road lights have to be lit, all the barriers have to be proved fully lowered, and the operator has to confirm that the crossing is clear by pressing a crossing clear button.

As soon as a protecting signal shows a proceed aspect (or tries to), approach/route locking is activated (*). This now locks the crossing controls. The operator now will be unable to raise the barriers.

The approach/route locking will remain effective until either, the interlocking sees the correct sequence of track circuits (or axle counters) being operated (that shows that the train has passed the protecting signal, in other words, normal passage of the train), the signaller cancels the signal route and the interlocking can see via the track circuits (or ACs) that there is no train in section or approaching (comprehensive approach locking)(if provided), or the route has been cancelled and the protecting signal is now red and then after a time delay (normally 2 or 3 minutes).

After the train passes the protecting signal, and until it is clear of the crossing, track circuits (or ACs) directly lock the crossing controls.

Normally, the barriers can be raised and the crossing opened to road traffic as soon as the train has cleared the track circuit (or AC) that is over the crossing assuming that the switch/lever/control for the signal has been restored to the normal (on/danger) position. You do not have to wait until the train has cleared the complete signal section.

The above is the general principles for unidirectional lines. It’s more complex for single lines, bidirectional lines or if points/junctions are involved.

* Note, actual practice has varied slightly over the years, there are many differences between different level crossing controls. Hence some have separate route locking and approach locking, while others combine it together. The end result is the same, the crossing controls are locked as described above.

Aha excellent thank you

So yes the behaviour I observed, the delay between the train clearing the crossing and the barriers raising despite being able to see the signaller stood by the controls, was likely due to the train not being clear of the track circuit that covers the crossing.

 

[Ken H]

How does it all work at a gate box. How is he informed he needs to close the gates, how does he tell the signallers the gates are closed, and how does he know to open the gates. What if there is a second train coming?

 

[MadMac]

How does it all work at a gate box. How is he informed he needs to close the gates, how does he tell the signallers the gates are closed, and how does he know to open the gates. What if there is a second train coming?

Differs from place to place. Typically, they’ll get some sort of “Train Approaching“ indicator which Local Instructions will tell the Crossing Keeper are the “cue” to close the crossing. In some cases, the signals over the crossing will be “slotted” by the crossing (signalling centre sets the route but the signals don’t clear unless the Crossing Keeper “releases“ them): Camps Junction was, for me, a rather bizarre one. The signals over the crossing had Edinburgh numbering, their phones went to Edinburgh, but they were entirely under the control of Camps. They had “Train Approaching“ indications for Down, Up (Stopping at Kirknewton) and Up (non-stop). One of the Crossing Keepers pointed out to me that had he followed the “letter of the law”, he could have the gates closed for 54 minutes at certain times!

 

[MarkyT]

That certainly would be correct now, but prior to the Moreton-on-Lugg incident that didn't apply at my Box, or others near-by. I had to prove to an RAIB inspector what we could do when he visited my Box in the week after that, and it was why my Box and others were fitted with treadles to lock the barriers until a train had passed, or to give us 5 minutes "thinking time" if we did want to raise the barriers for any reason if one hadn't.

After Moreton-on-Lugg ISTR there was a project to identify similar loopholes at other mechanical boxes and apply suitable new safeguards.

 

[Llanigraham]

How does it all work at a gate box. How is he informed he needs to close the gates, how does he tell the signallers the gates are closed, and how does he know to open the gates. What if there is a second train coming?

Depends what you mean by a "gate box". In AB areas that Box overlooking a LX is just as likely to be controlling a section and therefore will have it's own signals. Certainly that applies to most of the Crossing Boxes between Tram Inn and Marshbrook on The Maches line.
At Bromfield we started the closure actions on the DOWN when we received the bell code fron Onibury that a train had left Craven Arms and on the UP from the annunciator activation as a train left Ludlow Station. It made no difference if there were trains in both directions.

After Moreton-on-Lugg ISTR there was a project to identify similar loopholes at other mechanical boxes and apply suitable new safeguards.

Didn't I know it!!
I had one of the Inspectors turn up un-announced one afternoon (*) who wanted to watch everything I did and very closely questioned me about how and why we did things. I was lucky as the guy who did most of my Box training knew that we could do exactly what Adrian ended up doing at Moreton, and it was thoroughly thrashed into me to always check and if neccessary stick a collar on the lever to remind you. That was something I often did if things were getting stressful for some reason.

(*) I refused to let him in until he provided ID and even then made him sign the TRB. Got a "commendation" later as being the only Box that had done so!

 

[Annetts key]

That certainly would be correct now, but prior to the Moreton-on-Lugg incident that didn't apply at my Box, or others near-by. I had to prove to an RAIB inspector what we could do when he visited my Box in the week after that, and it was why my Box and others were fitted with treadles to lock the barriers until a train had passed, or to give us 5 minutes "thinking time" if we did want to raise the barriers for any reason if one hadn't.

That does not surprise me. It depends on when the interlocking and level crossing was designed and installed, and if/when any alterations took place. As problems with existing designs are found, the standard gets updated so it applies to new installations. However, retrospective work on existing installations, unless considered to be significantly important, will often not take place.

After Moreton-on-Lugg ISTR there was a project to identify similar loopholes at other mechanical boxes and apply suitable new safeguards.

I wonder if this was known about by BR signal engineers. As certainly I know of a mechanical signal box controlling a MCB CCTV crossing dating from 1973 (this is the date that the crossing was changed from a gated crossing to living barriers). This crossing definitely had approach controls, as a S&T maintenance team had to adjust the wire for one of the mechanical semaphore signals that protected it. While testing the signal operation, the signaller had to close the crossing. Unfortunately soon afterwards an ambulance turned up wanting to cross the crossing. It had to wait for the two minute timer (part of the approach locking) to run before the signaller could raise the barriers and let it and all the other road traffic through…

I should add that this crossing did have track circuits. There were also some treadles, but I can’t remember what function the treadles performed.

 

[edwin_m]

From the RAIB report into Moreton-on-Lugg: R042011_110228_Moreton_on_Lugg_v2.pdf (publishing.service.gov.uk)

174 A further causal factor was that although an engineered safeguard was provided in the form of interlocked signals, this was not sufficient to prevent a signaller mistakenly replacing the protecting signal and then raising the barriers when a train was closely approaching. Approach locking, an engineered safeguard that would have provided this protection, was not fitted when the crossing was converted to manual barrier operation in the mid-1970s. Although there have been a number of other incidents involving errors made by signallers and level crossing keepers in recent years, Network Rail has neither subsequently fitted it nor undertaken a formal risk assessment to quantify the safety benefit

...

178 Network Rail are planning to action two recommendations from the Formal Investigation it has led into the accident at Moreton-on-Lugg (paragraph 10): l a review of improvements to interlocking arrangements to mitigate the risk of operator error at MCB level crossings in semaphore signal areas.

...

179 A working group has been set up to implement the first recommendation, and work has been undertaken to determine the status of engineering safeguards fitted to other level crossings. The review has been expanded to cover all level crossings with interlocked protecting signals (not only in semaphore signal areas). It has identified that, including Moreton-on-Lugg, there are 54 MCB-type level crossings without approach locking, or with only partial protection.