ATP, TPWS and AWS

[gwr4090]

Am I correct in assuming that AWS is quite independent of TPWS and that
the two systems operate side by side ? On those lines/trains which are
also ATP fitted (eg GWML) are all three systems in operation
simultaneously, or is TPWS and/or AWS switched off ?

David

[David Gould]

Post by gwr4090
Am I correct in assuming that AWS is quite independent of TPWS and that
the two systems operate side by side ? On those lines/trains which are
also ATP fitted (eg GWML) are all three systems in operation
simultaneously, or is TPWS and/or AWS switched off ?
David

AWS runs in conjunction with ATP, as both were unavailable during that
unfortunate incident in 1997.

Further, AWS is used in conjunction with TPWS. TPWS only protects
signals near to junctions or other areas where there are a large number
of SPADs, where there would be a reasonable probabilty of a collision
were TPWS not fitted (by reasonable, I mean something of the order of
1:1000000).

I do not believe, however, that ATP and TPWS run concurrently, but the
both the relevant section of the GWML and the GW HSTs are fitted with
all three systems since only the Paddington-Bristol-Cardiff route is ATP
fitted - other routes out to Cornwall, for example, have received the
usual TPWS attention, however.

Further, Networkers do not have ATP equipment but do have TPWS, so TPWS
is needed also.

Not too sure about the 332s that run to Heathrow, however. They are ATP
fitted, as is the little branch line to Heathrow itself.

D.

[Roger H. Bennett]

Post by David Gould
AWS runs in conjunction with ATP, as both were unavailable during that
unfortunate incident in 1997.

I don't think that bit's right. IIRC the situation was that the AWS had
failed and the driver hadn't been trained on ATP so was not allowed to
switch it on.

Roger

[Brimstone]

Post by Roger H. Bennett

Post by David Gould
AWS runs in conjunction with ATP, as both were unavailable during
that unfortunate incident in 1997.

I don't think that bit's right. IIRC the situation was that the AWS
had failed and the driver hadn't been trained on ATP so was not
allowed to switch it on.

Which means it wasn't available.

[David Hansen]

On Sat, 30 Oct 2004 12:33:55 +0000 (UTC) someone who may be

Post by Brimstone
Which means it wasn't available.

Yes, but not for the reason postulated.

[Brimstone]

Post by David Hansen
On Sat, 30 Oct 2004 12:33:55 +0000 (UTC) someone who may be

Post by Brimstone
Which means it wasn't available.

Yes, but not for the reason postulated.

In this specific instance is the reason for that non-availability important?


(I'm sure most people on this ng are sufficiently aware of the various
failings surrounding that incident that they don't need to be repeated.)

[David Gould]

Post by Brimstone

Post by David Hansen
On Sat, 30 Oct 2004 12:33:55 +0000 (UTC) someone who may be

Post by Brimstone
Which means it wasn't available.

Yes, but not for the reason postulated.

In this specific instance is the reason for that non-availability important?
(I'm sure most people on this ng are sufficiently aware of the various
failings surrounding that incident that they don't need to be repeated.)

Erm, I honestly couldn't remember quite why (but clearly now no), and as
was mentioned, either way, it was not available.

Picky...

[Ross]

On Sat, 30 Oct 2004 12:16:21 +0100, David Gould wrote in

Post by David Gould
AWS runs in conjunction with ATP, as both were unavailable during that
unfortunate incident in 1997.

Not true. ATP could have been used without AWS on the GWT HST.

Neither driver of that train was currently competent with ATP [1] so
it hadn't been turned on.

Had the second been competent, the rules then in force did not allow
ATP to be switched in en route because it took 2-4 minutes to do so.

That refresher training and two minute delay would have saved a number
of lives... :(



[1] i.e., whilst both drivers had received initial training, they had
not used ATP after that training and had not then received the
necessary refresher training to enable them to use the system.

[Graeme Wall]

Post by Ross
On Sat, 30 Oct 2004 12:16:21 +0100, David Gould wrote in

Post by David Gould
AWS runs in conjunction with ATP, as both were unavailable during that
unfortunate incident in 1997.

Not true. ATP could have been used without AWS on the GWT HST.
Neither driver of that train was currently competent with ATP [1] so
it hadn't been turned on.
Had the second been competent, the rules then in force did not allow
ATP to be switched in en route because it took 2-4 minutes to do so.
That refresher training and two minute delay would have saved a number
of lives... :(
[1] i.e., whilst both drivers had received initial training, they had
not used ATP after that training and had not then received the
necessary refresher training to enable them to use the system.

Something that has always puzzled me about that incident, just how much
training do you need to operate a train with ATP enabled?

[Ross]

On Sat, 30 Oct 2004 19:17:30 +0100, Graeme Wall wrote in
<ddda1a64d%***@greywall.demon.co.uk>, seen in uk.railway:
[...]

Post by Graeme Wall
Something that has always puzzled me about that incident, just how much
training do you need to operate a train with ATP enabled?

I know nothing about ATP, so I don't know how complex it is to set up.

If a self-test takes anything up to 4 minutes, though, I can't see it
being that simple!

[David Hansen]

On Sun, 31 Oct 2004 22:32:15 +0000 someone who may be Ross

Post by Ross
I know nothing about ATP, so I don't know how complex it is to set up.

In the case of a HST the train is a standard length and has a
standard maximum speed and weight, so set up should be easier than
on a locomotive.

As far as operating a train once it is switched on, I think the
problem is placing the display into a modified standard speedometer.
This limits the display drastically. In particular the three segment
display takes a little time to understand, the green and yellow LEDs
are a bit easier. Undoubtedly it is not a problem once people have
used it a few times, but the problem at Southall was that both
drivers had not got to this stage (through no fault of their own).

The "wiggly wire" display may have been too complicated, but it was
well laid out for the time. I hope that is the sort of thing that
forms the basis for ETCS displays, rather than current ATP displays
(which tend to be basic on most railways as it was added in rather
than thought out from scratch).

[0.7ohm Drop Shunt]

I cant comment on ATP, but AWS and TPWS are completely seperate.

They are similar in what they are attempting to do, but very much
different in practice.

The AWS is Fail Safe - the Permanent Magnet being able to work no
matter if the power has failed. The Electro is energised when the
signal is Green.

TPWS on the other hand, isnt fail safe. Take it away, and the driver
doesnt know any different.
The Loops are only active when the signal is at danger,.
In the case of ones to prevent overspeeds into a PSR or Platform End,
they are energised all the time, but if they fail, wont make a jot of
difference.

[David Hansen]

On 30 Oct 2004 04:22:11 -0700 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
TPWS on the other hand, isnt fail safe. Take it away, and the driver
doesnt know any different.

Take the loops away when they are energised and the signal in the
rear will be replaced or maintained at danger, because TPWS is wired
into the lamp failure part of the signalling system. This gives
automatic double blocking at the locations covered by TPWS. This is
one of the reasons the cost of TPWS went up from the initial
estimates. I am agnostic on whether it was worth the money.

Post by 0.7ohm Drop Shunt
In the case of ones to prevent overspeeds into a PSR or Platform End,
they are energised all the time, but if they fail, wont make a jot of
difference.

I assume their failure is detected and indicated in some way, but I
have never seen this mentioned in descriptions.

[0.7ohm Drop Shunt]

Post by David Hansen
On 30 Oct 2004 04:22:11 -0700 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
TPWS on the other hand, isnt fail safe. Take it away, and the driver
doesnt know any different.

Take the loops away when they are energised and the signal in the
rear will be replaced or maintained at danger, because TPWS is wired
into the lamp failure part of the signalling system. This gives
automatic double blocking at the locations covered by TPWS. This is
one of the reasons the cost of TPWS went up from the initial
estimates. I am agnostic on whether it was worth the money.

Most of the TPWS that I have is on a mechanical railway, and is wired
into the system through additional conboxes under the frame.
In a recent failure, where the power supplying all the TPWS equipment
went down, the TPWS was inoperative. All that happened was the
indication in the box went to the failed position, and the signaller
carried on as normal.

No modules supplied, no loops energised, no double blocking. nada.

Post by David Hansen

Post by 0.7ohm Drop Shunt
In the case of ones to prevent overspeeds into a PSR or Platform End,
they are energised all the time, but if they fail, wont make a jot of
difference.

I assume their failure is detected and indicated in some way, but I
have never seen this mentioned in descriptions.

No indications whatsoever for the ones I have.

[David Hansen]

On 30 Oct 2004 09:37:41 -0700 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
Most of the TPWS that I have is on a mechanical railway, and is wired
into the system through additional conboxes under the frame.
In a recent failure, where the power supplying all the TPWS equipment
went down, the TPWS was inoperative. All that happened was the
indication in the box went to the failed position, and the signaller
carried on as normal.

Presumably a training/rules problem.

Post by 0.7ohm Drop Shunt

Post by David Hansen
I assume their failure is detected and indicated in some way, but I
have never seen this mentioned in descriptions.

No indications whatsoever for the ones I have.

Anyone know what the position is in power box areas?

[0.7ohm Drop Shunt]

Post by David Hansen
On 30 Oct 2004 09:37:41 -0700 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
Most of the TPWS that I have is on a mechanical railway, and is wired
into the system through additional conboxes under the frame.
In a recent failure, where the power supplying all the TPWS

equipment

Post by David Hansen

Post by 0.7ohm Drop Shunt
went down, the TPWS was inoperative. All that happened was the
indication in the box went to the failed position, and the signaller
carried on as normal.

Presumably a training/rules problem.

Nope, its the way it is. Do you seriously want train delays for an
Additional System?

Post by David Hansen

Post by 0.7ohm Drop Shunt

Post by David Hansen
I assume their failure is detected and indicated in some way, but I
have never seen this mentioned in descriptions.

No indications whatsoever for the ones I have.

Anyone know what the position is in power box areas?

That does include the PSR's that I have in my PSB area.

[David Hansen]

On 31 Oct 2004 04:39:46 -0800 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
Nope, its the way it is. Do you seriously want train delays for an
Additional System?

What I want is immaterial. Those in charge have decided that failure
of the loop in colour light areas introduces double blocking. I
assumed something similar was part of the rules in semaphore areas.

As this is not the case I might ask why.

[Wobbly Bob]

Post by David Hansen
On 31 Oct 2004 04:39:46 -0800 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
Nope, its the way it is. Do you seriously want train delays for an
Additional System?

What I want is immaterial. Those in charge have decided that failure
of the loop in colour light areas introduces double blocking. I
assumed something similar was part of the rules in semaphore areas.
As this is not the case I might ask why.

TPWS isn't intended to guard against tail end collisions, that is, it
isn't fitted if that is the only risk. For this reason, taking a box
with Home and Section signals with converging pointwork in between
them, TPWS will be fitted at the Home (to protect the conflict), but
not at the Section signal.

I understand the original intention was that the equipment would be
monitored (in all *signalling* cases) by the filament failure circuit,
a non-vital indication which does not affect the signal in rear.
This is broadly parallel with the level of monitoring provided on
semaphore signals. Somewhere along the line, someone decided this
wasn't enough and it was changed to prove it in the lamp repeating
circuit instead. Of course, there isn't a lamp proving circuit on
(most) semaphore signals as the lamp is not critical to the signal
indication.

Wobbly Bob

[David Hansen]

On Sun, 31 Oct 2004 15:25:54 +0000 someone who may be Wobbly Bob

Post by Wobbly Bob
TPWS isn't intended to guard against tail end collisions, that is, it
isn't fitted if that is the only risk.

Here is a modern local variation:-)

On at least the Edinburgh & Glasgow line, as well as that part of
the ECML through Fife which is controlled from Edinburgh power box,
TPWS is fitted to signals in the rear of station platforms. This can
only be to prevent or mitigate rear end collisions. This strikes me
as a very sensible and cheap use of TPWS. It certainly provides more
protection to passengers than TPWS at buffer stops.

Why this has been done I cannot say and have never asked. Perhaps it
is a collective memory of what was the most devastating peacetime
collision in Britain when it happened, Castlecary.

Post by Wobbly Bob
Somewhere along the line, someone decided this
wasn't enough and it was changed to prove it in the lamp repeating
circuit instead. Of course, there isn't a lamp proving circuit on
(most) semaphore signals as the lamp is not critical to the signal
indication.

Thanks, that makes sense.

[Wobbly Bob]

Post by David Hansen
On Sun, 31 Oct 2004 15:25:54 +0000 someone who may be Wobbly Bob

Post by Wobbly Bob
TPWS isn't intended to guard against tail end collisions, that is, it
isn't fitted if that is the only risk.

Here is a modern local variation:-)
On at least the Edinburgh & Glasgow line, as well as that part of
the ECML through Fife which is controlled from Edinburgh power box,
TPWS is fitted to signals in the rear of station platforms. This can
only be to prevent or mitigate rear end collisions. This strikes me
as a very sensible and cheap use of TPWS. It certainly provides more
protection to passengers than TPWS at buffer stops.
Why this has been done I cannot say and have never asked. Perhaps it
is a collective memory of what was the most devastating peacetime
collision in Britain when it happened, Castlecary.

It's probably because, with a station ahead, there is a higher than
average chance of the signal being at Red when encountered by a train.
This is a factor in risk analysis for signals (as is a station in
rear of the signal), but I'm not aware that it is, on its own, a
reason for fitment. As you say, possibly a local variation, or I'm a
bit out of date on the principles applied - whilst the high profile
TPWS programme is complete, at least in its basic form, its
application is still being developed to improve its effectiveness at
higher speeds.


Wobbly Bob

[0.7ohm Drop Shunt]

Post by Wobbly Bob

Post by David Hansen
On 31 Oct 2004 04:39:46 -0800 someone who may be "0.7ohm Drop Shunt"

Post by 0.7ohm Drop Shunt
Nope, its the way it is. Do you seriously want train delays for an
Additional System?

What I want is immaterial. Those in charge have decided that

failure

Post by Wobbly Bob

Post by David Hansen
of the loop in colour light areas introduces double blocking. I
assumed something similar was part of the rules in semaphore areas.
As this is not the case I might ask why.

TPWS isn't intended to guard against tail end collisions, that is, it
isn't fitted if that is the only risk. For this reason, taking a box
with Home and Section signals with converging pointwork in between
them, TPWS will be fitted at the Home (to protect the conflict), but
not at the Section signal.
I understand the original intention was that the equipment would be
monitored (in all *signalling* cases) by the filament failure

circuit,

Post by Wobbly Bob
a non-vital indication which does not affect the signal in rear.
This is broadly parallel with the level of monitoring provided on
semaphore signals. Somewhere along the line, someone decided this
wasn't enough and it was changed to prove it in the lamp repeating
circuit instead. Of course, there isn't a lamp proving circuit on
(most) semaphore signals as the lamp is not critical to the signal
indication.
Wobbly Bob

Exactly, and in the case of semaphore signalling, TPWS is not fail safe
as in the way that AWS is.
It is protected in a total failure mode by the Rule Book, whereas all
other equipment failing is protected by the system itself.

The actual instructions for a TPWS failure, which will not activate a
brake application, *and the signal can be given* is to not alert the
driver. However, if the signal cannot be given the driver will be
advised.

[David Hansen]

This post might be inappropriate. Click to display it.

[RPM]

Post by gwr4090
Am I correct in assuming that AWS is quite independent of TPWS and that
the two systems operate side by side ? On those lines/trains which are
also ATP fitted (eg GWML) are all three systems in operation
simultaneously, or is TPWS and/or AWS switched off ?

Where ATP is in use, AWS and TPWS remain active. The difference on ATP
fitted infrastructure is that if AWS/TPWS fails, the train can continue at
line speed and the usual Rulebook requirements do not apply.

Roger

http://rpm-railpics.fotopic.net/
http://therailwaystationgallery.fotopic.net/
http://therailticketgallery.fotopic.net/