SSR signalling contract to be re-let (again)

Just another thing to add from the Barons Court Sub Gaps going west the traction supply will remain at 630v and the regenerative braking will automatically be disabled on the S stocks.

Quick update : First S stock fitted with Thales signalling has started dynamic (i.e. moving rather than stationary) testing on the test track at Old Dalby (now known as RIDC Melton by Network Rail)

Do you know whether L29 is undertaking dynamic testing now as well?

L29 has certainly operated on the test track several times over the last month; my understanding is that this isn't (yet) for ATC purposes.

The S stock at Old Dalby is train 142 - i.e. units 21467/8, which has yet to be delivered to London.

I assume that the Richmond branch will be the same as the Wimbledon branch.

But I think that there is a better solution. Quite simply at Acton Town, Gunnersbury and Putney Bridge (or East Putney) the trains should switch between automated and human driven modes.

In other words, train services to the three western District Line termini remain human driven. 

There are several benefits from this.

1) Helps the train drivers retain their safety certification which would otherwise lapse if they did not drive trains at least every 6(?) weeks.

2) At a time of severe financial stress* this will save costs. Since there will be no operational benefit (ie; no increase in line capacity) and the signalling on these routes is assumed to be in good nick so not spending money on them could help keep the project alive.

If it were not for what I understand to be an urgent need to replace life expired signalling I would even suggest that to save money the automation should not be installed west of Hammersmith or north of Harrow-on-the-Hill (neither FL or SL). 

But I would still automate all the way to Upminster and Uxbridge, as well as the Circle Line routes, as these areas will benefit from much needed increases in train frequencies. Since the Olympia branch is so close to the central area it probably should end up being automated, as otherwise trains would need to switch modes at Earls Court and this might cause delays at a very busy station where trains should never stop for longer than absolutely necessary.

*As its OT for this thread all I'll say is that my reason for suggesting such drastic changes to whats planned to lower costs and thereby save the entire project is that I am expecting another financial 2008 event. More severe. Possibly as soon as spring 2016. The resignalling with automation is urgently needed, it would a terrible shame were matters financial to cause it to be delayed or cancelled. 

Simon 

That is exactly how its going to work

On Richmond and Wimbledon the trains will still be driven in ATO but otherwise it is as you describe.

For Simon - looking at saving money is one thing as is the balance of capacity vs cost however there are other factors. For example, if the trains are running under ATO throughout the Met line they will be regulated automatically meaning that the headways will be more even and trains will arrive at Harrow more regularly improving junction working. As another example using the legacy signalling will not bring improvements in customer information (train arrivals) and would lead to some cabins still being staffed and Hammersmith control centre being under-utilised. All of these factors, and many more, have been considered and weighed up when seeking financial authority for the project last year.

My understanding was that the 73 stock was due to have a replacement speedo with a TOD style display, relevant signalling hardware and would run in PM while on Uxbridge branch, has this changed?

I am somewhat baffled here as the information about the Wimbledon and Richmond branches in the two messages above seem to contradict each other.

Or is it that there will be a new quasi-ATO mode designed for routes where there are no radio signals; so that despite not being in constant contact with the central computers the trains are fully functional (including not speed limited) and instead the human in the cab at the front of the trains drives normally... obeying traditional lineside signals?

(This being in addition to the manual modes used on other ATO lines on Sundays where the train is still in contact with the signalling computers but to retain safety certification the human driver is actually driving) 

-------------------

Another question relates to Barons Court - Acton. Will the District and Piccadilly be restricted to their own tracks (as per Finchley Rd - Wembley Park) or retain the freedom to divert, if required?

Simon

Can trains be operated in Coded Manual with Seltrac or would drivers be restricted to manual at reduced speed.

PM on Seltrac is pretty rubbish, due to a combination of poor design, plenty of pitfalls where the driver can be lulled into an error which might cause a delay, and a massive number of bugs in the system - although to be fair there has recently been a push to fix some of these. Worst of all, the system is still set to the minimum brake rate in the open sections, so you are dumbed down to ATO levels of performance which can't tell whether adhesion is likely to be bad so assumes the worst case scenario (because Seltrac has a design flaw which sends the train non-communicating if the wheels slide).

All this said, with practice it's perfectly possible to drive charismatically in PM, and it's quite possible to maintain or beat the ATO running times, because Thales/Seltrac ATO isn't particularly good either.

There are a handful of drivers on the Northern who regularly drive in PM, and through practice and experience have become excellent - they maintain time and never have issues. Unfortunately by contrast some drivers haven't picked up PM too well and avoid it as far as possible.

I know why the NCT issue occurs, but it's hardly a desirable feature that a train is likely to go NCT if a wheelslide occurs, after all this is hardly a rare occurrance in train operations. Certainly the Central Line doesn't suffer from this weakness. I'd say it's a pretty major design flaw if you can't brake heavily in the open-air without risking causing a (potentially lengthy) delay.

Comparison with the legacy signalling on the Jubilee is not a fair comparison. On the Stanmore-Charing Cross section most of the signalling dated from the 1980s, and was designed at a time of falling traffic, it was never specified, designed or intended to provide an intense service. The signalling on the JLE was a temporary fix, designed to match the capability of the rest of the line, and constrained by incorporating elements of the abortive Westinghouse moving block system.

It's worth mentioning that the Charing Cross branch operated a frequency of 44tph long ago, compared to that benchmark TBTC is hardly impressive. Unlike the Victoria Line where DTGR is achieving 34tph on a daily basis, we wait to see whether Seltrac will be able to reliably match this, I have a suspicion it may prove optimistic.

What with the wasted time and money, and the disastrous flirtation with CityFlo, and now being boxed into a position where Thales is the only supplier on the table, I think in hindsight LU will come to regret the decision to dump the original Westinghouse SSR contract.

I thought the ditching of Westinghouse was one of the first things LU did when they got hold of Metronet.

Regarding the NCTs, the whole problem is that trains *don't* brake heavily in the open, because the system is set to the minimum brake rate. Any heavy braking is basically likely to produce a NCT if the railhead is less than perfect.

The use of low brake rates gives needlessly extended journey times, which is bad for the passenger. It's just about tolerable on the Northern as only a relatively small amount of the line is in the open, and the effect has been a little masked by the fact the line previously had a low line-speed compared to other lines. However the Jubilee from Stanmore to Finchley Road is a painful experience, and a journey like Upminster to the centre will become probably unacceptably long if a solution isn't found to this issue. And of course there are numerous reasons why a train might need to emergency brake, eg a hazard on the line ahead or a PEA pulled leaving a station - these everyday scenarios shouldn't cause the delay of an NCT.

Personally, observing the routing of trains via TBTC, how trains follow each other at certain locations, and observing the working at key junctions, I have doubts the Northern will be able to reliably run much more than it does at present. Platform re-occupation times at some key locations are pretty poor as things stand at present. The glitzy promo videos from Thales and Alcatel before them are somewhat different to reality.

So at 44tph can good 'ole human driving (with the right signalling which I assume included speed control when approaching stations and junctions) ever be bettered by artificial intelligence TBTC signalling?

Maybe though in the old days there were fewer safety systems, such as a requirement for a train to be fully stationary before the doors are opened. This alone adds a few seconds to station stop durations.

Possibly its unfair to compare the NCT issue with simple train detection via track circuits or axle counters as these too can experience foibles.

Perhaps though the radio communications system will be more resilient to trains which experience wheel slip than the leaky wires used on the Jubilee Line / Northern Line / DLR? Otherwise journey times to Amersham and Upminster will potentially end up taking more rather than less time... and those passengers who can will switch to other services.

Simon