When (OK if) we get to ETCS level 3 it will feature a "moving block".

On a conventionally signalled railway and ETCS level two each block is
a fixed length and it doesn't matter if the train occupying the block
is a "153" or a 66 at the head of a long line of container flats.

With ETCS level three is the "block" around the train the same size
irrespective of the actual train or would the block for a 153 be
smaller than "my" 66+container flats?

I suspect relative speeds come into the picture somewhere but at the
moment I am simply interested in the block size.

On 27.07.23 22:37, Graham Harrison wrote:
> When (OK if) we get to ETCS level 3 it will feature a "moving block".
>
> On a conventionally signalled railway and ETCS level two each block is
> a fixed length and it doesn't matter if the train occupying the block
> is a "153" or a 66 at the head of a long line of container flats.
>
> With ETCS level three is the "block" around the train the same size
> irrespective of the actual train or would the block for a 153 be
> smaller than "my" 66+container flats?
>
> I suspect relative speeds come into the picture somewhere but at the
> moment I am simply interested in the block size.

What is needed is for the system logic to be arranged in a way that
permits a single train to occupy multiple blocks. This can be done with
conventional signalling systems too. For example there have been
stations on the London Underground for a long time that have platforms
divided into two blocks so that an arriving train can enter the station
before the departing train has fully cleared the platform. In this
situation, a train standing at the platform will be occupying both
blocks simultaneously. It's not really all that different in principle
from the old fashionned absolute block signalling system where a box can
be closed by effectively linkign the two boxes either side and combining
the two blocks into one long one, but obviously computers can do it
faster and more efficiently.

Robin

Am 28.07.2023 um 08:51 schrieb Bob:
> On 27.07.23 22:37, Graham Harrison wrote:
>> When (OK if) we get to ETCS level 3 it will feature a "moving block".
>>
>> On a conventionally signalled railway and ETCS level two each block is
>> a fixed length and it doesn't matter if the train occupying the block
>> is a "153" or a 66 at the head of a long line of container flats.
>>
>> With ETCS level three is the "block" around the train the same size
>> irrespective of the actual train or would the block for a 153 be
>> smaller than "my" 66+container flats?
>>
>> I suspect relative speeds come into the picture somewhere but at the
>> moment I am simply interested in the block size.

In the German system
<https://de.wikipedia.org/wiki/CIR-ELKE>
virtual Block size is 300m for high-speed lines, down to 50m for S-Bahn
Munich in station areas.

Relative speeds do not come into the picture, we're still operating
"entire braking distance must be safe now" rather than "will be safe by
the time reach the point". The speed-dependent aspect is ability to
brake, taking into account topography (braking uphill is significantly
faster than downhill).

Rolf

Am 28.07.2023 um 10:52 schrieb Rolf Mantel:
> Am 28.07.2023 um 08:51 schrieb Bob:
>> On 27.07.23 22:37, Graham Harrison wrote:
>>> When (OK if) we get to ETCS level 3 it will feature a "moving block".
>>>
>>> On a conventionally signalled railway and ETCS level two each block is
>>> a fixed length and it doesn't matter if the train occupying the block
>>> is a "153" or a 66 at the head of a long line of container flats.
>>>
>>> With ETCS level three is the "block" around the train the same size
>>> irrespective of the actual train or would the block for a 153 be
>>> smaller than "my" 66+container flats?
>>>
>>> I suspect relative speeds come into the picture somewhere but at the
>>> moment I am simply interested in the block size.
>
> In the German system
> <https://de.wikipedia.org/wiki/CIR-ELKE>
> virtual Block size is 300m for high-speed lines, down to 50m for S-Bahn
> Munich in station areas.
>
> Relative speeds do not come into the picture, we're still operating
> "entire braking distance must be safe now" rather than "will be safe by
> the time reach the point".  The speed-dependent aspect is ability to
> brake, taking into account topography (braking uphill is significantly
> faster than downhill).

Reading further, it seems that we need some clarifications what exactly
we speak about: what is a 'block'?

A traditional (point-wise) block consists of three aspects:
1) Safe Location Determination of a train (i.e. the signal box knows
there a train head is)
2) a signal (free/ halt and speed limits) - speed limits typically state
the minimum speed allowed anywhere in the block
3) a device to detect 'block empty', e.g. a wheel counter.

Note: constant block size is *not* an essential aspect of traditional
block but it greatly simplifies signal box logic ('half block mode'
where braking distance is longer than the block size).

Simple In-cab signalling moves the signal safely from fixed points to
the cab (signals might continue to exist line-side as a fallback
mechanism or for mixed mode), the line-side logic is still traditional.
In-cap signalling

Next step is 'continuous block' (German LZB, CIR-ELKE as above, ETCS
level 2). Location determination of the train is extended to high
resolution; typically as a mixture of line-side detectors (LZB: cables,
ETCS: Balises), speed measurement and safe communication between
lineside and the train. You reach capacity gains by enabling
high-quality speed limit calculations.
'Block empty' devices stay at the old block limits; the only necessary
action to reduce block size is adding a few more wheel counters at
critical locations.
In 'continuous block' mode, speed limit changes (especially increases)
can be signalled at any location; typically line-side signals (if they
still exist as fall-back) are switched off when their meaning would
deviate from the in-cab signal to minimize confusion to the driver.

The only aspect missing for 'moving block' of ETCS level 3 is the
frequent / continuous 'block empty' detection; apparently, there was a
recent decision to replace the idea of a static switch from ETCS level 2
to ETCS level 3 by an 'optional approach': the 'block' is free as soon
as either static wheel counter (level 2) or an safe train-end device
(level 3) declares the line to be free, so a train-end device can reduce
the gap to the following train by up to a minute on a non-optimized main
line (2km blocks at 120 km/h), hopefully giving the density advantages
of CIR-ELKE without the need for expensive line-side equipment.

Rolf