Kings Cross Signal Box Closed - Fire Alarm Activation

I was caught up in this this evening waiting for 1P10: ran about 30 mins late.

What I was wondering is what happens when an alarm goes off, are the train left to run to the next red signal as routed or do they get put back to Red? What happens when drivers start ringing in and there's no one there?

Peter

I was caught up in this this evening waiting for 1P10: ran about 30 mins late.

What I was wondering is what happens when an alarm goes off, are the train left to run to the next red signal as routed or do they get put back to Red? What happens when drivers start ringing in and there's no one there?

Peter

" said:

I was caught up in this this evening waiting for 1P10: ran about 30 mins late.

What I was wondering is what happens when an alarm goes off, are the train left to run to the next red signal as routed or do they get put back to Red? What happens when drivers start ringing in and there's no one there?

Peter

I was told last time I visited a large signalbox the usual procedure is to just let trains continue upto the next red signal. A GSMR broadcast is put out by control to advise drivers the box has been evacuated and to await instructions.

I hear that approach a lot from signallers but I must say I disagree. I always tell signallers to return signals to danger before evacuating. This is for the simple reason that on some areas of the network we have long sections of autos. So a train could potentially travel long distances without a signaller monitoring their progress. If an emergency were to arise there would be no one to manage it. while control could potentially put out a REC call on the GSMR on behalf of the absent signaller, a REC call isn't the only action a signaller would take in an emergency so having trains running around the network without a signaller monitoring their progress seems dangerous to me.

" said:

I hear that approach a lot from signallers but I must say I disagree. I always tell signallers to return signals to danger before evacuating. This is for the simple reason that on some areas of the network we have long sections of autos. So a train could potentially travel long distances without a signaller monitoring their progress. If an emergency were to arise there would be no one to manage it. while control could potentially put out a REC call on the GSMR on behalf of the absent signaller, a REC call isn't the only action a signaller would take in an emergency so having trains running around the network without a signaller monitoring their progress seems dangerous to me.

Well I suppose one way of looking at it, you're taught from a young age that generally speaking when a fire alarm goes off you get out as quickly as possible. Let's take Radyr signalbox as an example, it's a fairly large NX panel driving two Solid State Interlockings. There are no Signal Group Replacement Switches, only one ESOC per interlocking. Do you:

1) Leave the building as soon as the alarm sounds get control to put out a GSMR broadcast
2) Stay and pull up every route and pull the ERS on each auto then leave
3) Pull both ESOC buttons then leave (and deal with the fall out from the nuclear buttons)

I can think of a few Westcad workstations where you'd have a similar set of choices.

I'm not saying you are wrong by the way, just trying to promote some discussion.

" said:

I hear that approach a lot from signallers but I must say I disagree. I always tell signallers to return signals to danger before evacuating. This is for the simple reason that on some areas of the network we have long sections of autos. So a train could potentially travel long distances without a signaller monitoring their progress. If an emergency were to arise there would be no one to manage it. while control could potentially put out a REC call on the GSMR on behalf of the absent signaller, a REC call isn't the only action a signaller would take in an emergency so having trains running around the network without a signaller monitoring their progress seems dangerous to me.

Well I suppose one way of looking at it, you're taught from a young age that generally speaking when a fire alarm goes off you get out as quickly as possible. Let's take Radyr signalbox as an example, it's a fairly large NX panel driving two Solid State Interlockings. There are no Signal Group Replacement Switches, only one ESOC per interlocking. Do you:

1) Leave the building as soon as the alarm sounds get control to put out a GSMR broadcast
2) Stay and pull up every route and pull the ERS on each auto then leave
3) Pull both ESOC buttons then leave (and deal with the fall out from the nuclear buttons)

I can think of a few Westcad workstations where you'd have a similar set of choices.

I'm not saying you are wrong by the way, just trying to promote some discussion.

" said:

I hear that approach a lot from signallers but I must say I disagree. I always tell signallers to return signals to danger before evacuating. This is for the simple reason that on some areas of the network we have long sections of autos. So a train could potentially travel long distances without a signaller monitoring their progress. If an emergency were to arise there would be no one to manage it. while control could potentially put out a REC call on the GSMR on behalf of the absent signaller, a REC call isn't the only action a signaller would take in an emergency so having trains running around the network without a signaller monitoring their progress seems dangerous to me.

Well I suppose one way of looking at it, you're taught from a young age that generally speaking when a fire alarm goes off you get out as quickly as possible. Let's take Radyr signalbox as an example, it's a fairly large NX panel driving two Solid State Interlockings. There are no Signal Group Replacement Switches, only one ESOC per interlocking. Do you:

1) Leave the building as soon as the alarm sounds get control to put out a GSMR broadcast
2) Stay and pull up every route and pull the ERS on each auto then leave
3) Pull both ESOC buttons then leave (and deal with the fall out from the nuclear buttons)

I can think of a few Westcad workstations where you'd have a similar set of choices.

I'm not saying you are wrong by the way, just trying to promote some discussion.

For those that are interested, the "nuclear" option Karl describes is that when an ESOC (emergency signals on control - one big red button per interlocking) is pressed for several seconds, the power is cut to the interlocking which means everything trackside goes to its most restrictive state (points locked in their last position, signals at red). But also all logging is also stopped (since the interlocking is dead) which means a loss of information if something were to happen - not to mention trying to work out what is where when everything restarts as the train describer is going to be little help.

These days SGRC (signal group replacement controls) are used in preference, which are on-screen controls (or buttons on panels) to put back signals in a specific area, such as 1-2 controls for a large station. The interlocking is still running (and logging) though, so full tracking of trains still takes place. Not a rapid way to replace all signals but then neither is pressing an ESOC for each interlocking for 15 seconds!

Touch those buttons. Takes ages and lots of money to put back to normal apparently!

" said:

Touch those buttons. Takes ages and lots of money to put back to normal apparently!

Do NOT touch those buttons.

Some have paper flaps to avoid being inadvertently pressed, or are even in a box away from the workstations. If you "touched" one of those buttons and managed to cause a reset I suspect that would be the last of your 'box visits.

GW43125, the fear when the IECC was introduced in the 1980s was that a workstation would lose contact with the interlocking and the signaller would be running blind. Cutting the power is a simple, effective, if drastic way of providing that safety cord. As I say, since then there are other methods, not to mention more confidence in the systems.

I see your point and agree that 99.999% of the time option one on your list above would work without issue. Its the small risk of a safety incident that concerns me.

I think at the end of the day it has to come down to the specifics of the different boxes. Different LOMs have different approaches. The area I work in, a train could travel 18 miles on autos, travelling over several level crossings and bridges etc. Between the OMs and LOMs we have decided that it would be an unacceptable risk for trains to travel this distance without a signaller at the workstation. My main argument during this discussion was that with no one manning the local phone line and concentrator on the panel or workstation an emergency call could go unanswered. We have access to both SGRC and ESOC for every workstation so it wouldn't be a problem to operate these before leaving.

If it is ever a choice between safety and performance I would pick safety every time, and I would encourage any signaller to do the same.

[quote="dwaynedibley" post=83649]

Has anyone actually performed a risk assessment, because you might be imposing some fairly substantial costs on people based on gut feelings.

Has anyone actually performed a risk assessment, because you might be imposing some fairly substantial costs on people based on gut feelings.

" said:

Has anyone actually performed a risk assessment, because you might be imposing some fairly substantial costs on people based on gut feelings.

That sounds like management were I use to work. They stand in front of us every day preaching safety is our 1st priority, buuuutttt when you raise a control for an assessed hazard the next statement is usually "can't it be done cheaper" or just straight out "it costs too much"

So safety is our 1st priority until the cost is known

I'll need to confirm, but in the same vein I think mission critical operations centers in the US may have fairly conservative evacuation policies for essential employees. Friend told me that Amtrak moved its dispatch office from 30th St Station to its own building in Wilmington because every time there was a bomb scare it was awkward having the dispatchers not evacuate.

This is an image of a Signal Group Replacement soft key on a Westcad system. It can also be used by keyboard commands. With Westcad unlike IECC, there are no physical buttons to use, everything is on screen or via keyboard commands.

" said:

This is an image of a Signal Group Replacement soft key on a Westcad system. It can also be used by keyboard commands. With Westcad unlike IECC, there are no physical buttons to use, everything is on screen or via keyboard commands.

Some westcads have switches.

" said:

" said:

This is an image of a Signal Group Replacement soft key on a Westcad system. It can also be used by keyboard commands. With Westcad unlike IECC, there are no physical buttons to use, everything is on screen or via keyboard commands.

Some westcads have switches.

Where? Not saying you're wrong, just curious.