Great Western Coffee Shop

From the Chris Gibb report on Southern Rail (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/619795/chris-gibb-report-southern-rail.pdf)


Funny how you find support (and a keyword that well describes it) in odd places for something you've been looking at / even doing for years, isn't it?

TransWilts, from December 2013, has had "firebreaks" ... although originally for a different reason, which was to ensure that the peak evening carriage left Swindon at 17:36, rather that at 16:36 and 18:36.   But, yes, it has the effect of giving a bit of catchup time in the schedule / somewhat reduced (but justified) by the extension to Frome of one service. On a Saturday, the pre-2013 skeleton service was provided by trains which were (in essence) firebreaks in the Stroud Valley and Westbury area local services - and indeed they still run to this day, giving a welcome firebreak on the TransWilts carriage itself.

For our Melksham bus proposals, one area of concern is the robustness of road to rail connections at the station. We have extended the breaks in the middle of the day; the intent was originally to keep driving hours within legal limits, yet provide service in what should be peak hours, but the timetabling also has the very welcome effect of allowing things to get back on schedule if it's gone "tits up" during the morning.  See http://www.mbug.org.uk/bus_option_c.pdf for the full documentation and http://www.mbug.org.uk/m10.pdf for the overview of the plans.

So - THANK YOU, Chris Gibb, for using this word that so highlights what we're doing, and confirming it as a sensible idea from an independent source.   Is a similar technique in use elsewhere on GWR services?  Are they seen as a good idea fromt the customer angle?  Should there be more of them?

I find it difficult to answer the question. A service gap of say 30 minutes in a 15 min service interval would be OK.  A gap of 4 hours in a 2 hour service defiantly not. Somewhere between these extremes is a reasonable compromise.  It also depends on the amount of off peak traffic there is.

It sounds like you may be voting "only on some service groups" ... which is one of the options offered  ;D

I voted "no".
In general, in my view public transport should offer a regular service throughout the working day.

A mid afternoon shutdown might be JUST acceptable in a short term emergency, but regularly ? no way!

And from what I have seen of the way some TOCs conduct themselves, it might well make things worse. Examples could include having "early" shift staff work until about 14-00, and the late shift not starting until 16-00.
Trying to maintain track and trains in the afternoon gap and the work overrunning and disrupting the evening peak.

And in the case of GTR what about passengers making long journeys of the which the cross London GTR section is but a small part ?
It would also make our railways the laughing stock of the world.

If I firebreak is going to be added between 12 and 2, I wonder how effective it will be.  Any disruption starting after 1400 won't be mitigated.  Surely if there is disruption in the morning it will have sorted itself out by the start of the evening rush-hour in most cases.

I would agree with that ... but Chris Gibb was advocating a reduction of service for a period of hours in the middle of the day, and not a shutdown during mid afternoon.

Particularly in the context of recent tragic events, firebreak implied something completely different to me. Isn't there already a different term in use for this: overlay or something? Can't quite remember but it'll come to me.

As to the concept, a change from 10 to 15 minute intervals seems acceptable (chance would be a fine thing!) but a complete absence of services or introducing extra changes might have a knock on effect in putting off customers who normally travel at those times from trying the service at other times.

No.

Thin end of the wedge. We are used to regular service patterns and any reduction in service is a retrograde step.

Next it will be cuts after 8pm to prepare for overnight engineering, lengthening the midday firebreak from 1100-1500...

There probably is a different term ... Chris Gibb wrote his report back in December 2016 before those recent tragic events, so can be forgiven the term, but it wasn't released until it was the optimum time for those releasing it to do so.

The modelling I've been doing - on buses, where we're looking at running each bus with a single driver each day - comes out with some interesting results:

(http://www.wellho.net/pix/busfirebreak.jpg)

At present, the two town buses each stops for a legal minimum period at lunchtime, and both cover morning school runs and afternoon school runs, with a little bit afterwards too, but no coverage for peak commuters.   That and the fact they don't serve the station probably effects why they don't carry commuters  ;D

Proposal is to make the breaks in the middle of the day somewhat more generous, so that one vehicle can arrive early and do the morning commute, and the other stay until it has done the evening commute.   Middle of the day it reduces the 3 services per hour (at :12, :13, and :42 from Lowbourne) to 2 in the hour (at :13 and :42), with a 20 minute service at ENCTS peak time.  Interesting the see the journey number projections, which if realised would provide a significant extra income from the farebox, and would help us step away from the downward spiral these service have been on.


I hate to loose a regular pattern (not that we have one at lunch time even now on these buses), but I have the feeling that if we don't move the capacity to where it would be useful, the "regular" service will drop to there regularly being nothing.   Or should the subsidy be increased to provide an extra driver to carry 2 or 3 passengers per run at the quietest time of day?  I have no easy answer here ...

Although I'm in a completely different industry we just call it 'downtime' where resources and services are scheduled less or with gaps to allow maintenance or upgrades. As mentioned above though it does cause problems with scheduling and rotas.

I sort of assumed we already had it on GWR as from Twyford after 9.00 till early evening services are much reduced in comparison to 'rush hour'. I'm not against it in principle as fewer people travel at these times but it depends how much of a reduction it would mean. Also what would you actually do with the time? Move trains and crews around? Maintenance and repairs? Could you do much useful work in two hours?

There were a couple, of a fashion, introduced on the St Ives line last summer, where some services from St Ives were timetabled to be terminated at Lelant Saltings and not go on to St.Erth. This was to catch up with any delays caused by loading and unloading. Seemed to work ok

That does have a logic to it, as the timetable is so tight to enable a 30 minute service and I can well imagine in high season dwell times become a problem.

The introduction of a "fire break" in services is a DfT franchise decision,  TOC bid to run services based on requirements spec as part of the tendering process.

It is a "political" decision to inform for example the passengers at Iver, Langley, Burnham, Taplow (note just a for instance) that their half hourly service will become hourly during the day to "regulate" trains, this of course will apply in many places across the network, MP's will be asking questions.

The report also states the poor condition of the infrastructure on the "Southern" routes again DfT has to make the time available for NR to access the track and other equipment, the ORR is starting to realise this and are lobbying the DfT.  There are many pieces of equipment that was installed to N-1 so it could be taken out of service for maintenance, repair, replacement also to allow for it to fail with out effecting service; however there has been a policy by DfT for many years if not decades to "sweet the assets"

Despite the massive investment in the railways over the last 15 years, this investment has at best caught up with demand but demand is outstripping the investment.  The next 2 to 3 control periods (10 to 15 and potentially 20 years) will see Government investment in the National railway infrastructure scaled down in favour of HS2, the Government rhetoric however will say the opposite

Fascinating to read the views and thoughts.   

Where I grew up (North West Kent), we had longer distance commuter trains that did a morning run into London and then sat for the day and did an evening run out.   The entire fleet of Hastings Diesels overnighted in - err - Hastings ( Ore / St Leonard's / Bexhill) and travelled up to London in the morning peak.   One turned around at Tonbridge to provide the first southbound service; all the others carried on up to Charing Cross or Cannon Street, and then they went for the day to sidings at Grove Park that were empty overnight and at weekends, only coming out again in the late afternoon to do the homeward run. 

Same thing with Kent Coast expresses to Cannon Street parking up on the bridge at Blackfriars - 4CEP and 4BEP units just doing 1 round trip a day.

Sort of super-firebreak ... and hideously inefficient.   And I suspect you'll still see that sort of thing with Oxford HSTs into Paddington, and quite a number of the SWT 159s park up during the day too.

The bus scenario I look(ed) at somewhat differs; it's aimed at efficient crew breaks rather more than service robustness, mainly because (in our parts at least), the bus operation day is short enough to simply and cheapen operation with one driver per bus each day.

The other time saver I've seen there, is that there is also two drivers, one in each end so they don't have to walk the length of the two units along a crowded platform.

Actually there is not much difference regards the Kent services, units stabled at remote ends for the am peak services, the parked up just outside London during the day ready for the pm peak.

GWR seem to a lot of 165/6 moves to Reading after the morning peak and the a number of empties back to Padd for the evening peak.   There are some kept at OOC.

There is no easy solution to this, unless the general commuting population moves away from the "9 to 5" work pattern

There's another Network Rail usage of the term firebreak.   The peak fast line SWT timetable through Clapham Junction is determined on a 24 tph cycle, but they also leave a couple of paths empty every half hour or so.   The intention is that minor disruptions don't affect every successive service.  The gaps have been referred to in previous strategy documents such as the London and SE RUS as "firebreak paths", there are descriptions of how 28 tph could be achieved if the firebreak paths were filled, but the downside would be greater unreliability.

Paul

Chris Gibb proposed his "firebreak" in the specific context of Southern's complex timetable, and more particularly the crew and train diagramming:


In order to get all the pieces back into order after being jumbled up you need a bit of breathing space; either an off-peak timetable well below the peak capacity or his firebreak. Note that he only proposes this for a year to two, with the long-term solution being that reduction in off-peak service levels, and he also proposes to reduce the use of splitting/joining, relatively common on Southern, but notes:


Of course he is pushing in the opposite direction to DfT and TfL, who favour all-day metro service levels. Thus next year's new SWR timetable has 4 tph all day serving Reading and Windsor, but no extras in the peaks. While these are simpler end-to-end operations, complicated only by half the Reading trains going via each of Richmond and Hounslow, I'm sure a recovery strategy is needed and is being planned.

At the moment the usual approach is to cancel or part-cancel (by turning short) trains so as to slot trains back into paths. I'm not sure whether that always gets the train and crew into its original path; I suspect not, but for out and back running that's not a big issue. This turning short always happens at the outer end - I've never seen it done skipping Waterloo. Now I can guess why: crew turns start and end at Waterloo so it can't be skipped. But I have a hunch that crews don't often live near there, and travel in, so perhaps that could be altered.

The advantage would be that the cancelled trains would be over parts of the line where many other trains run, so each passenger's delay would be shorter. It might even be acceptable enough that the recovery break can be built into the timetable, so some interpeak trains will involve a change. Being timetabled has an advantage for passengers needing assistance, who won't call for an unexpected staff presence to help them, though of course would still happen during the disruption.

Crews still have to book on & off at their home 'depot'....they can't just alight from a late train as past their booking off time owing to lateness & go directly home!

Absolutely spot on, sounds like the start of a race to the bottom, and I am sure that the Unions would be up in arms at the inevitable threats to jobs that this would involve, we don't need even more needless strikes thanks!!!

Please try to understand that life is not 9-5 Mon-Fri any more, and if anything we should be looking to enhance services to reflect this, not cut them.

Let's take a station on the Southern (and SWT) network I once knew well, dear old Leatherdead. Throughout the day, there are six or occasionally seven trains per hour to London (http://www.brtimes.com/#!board?stn=LHD&date=20170626).

Chris Gibb defines a firebreak as "every route having at least one service missing from its hourly clockface pattern". Could Leatherhead cope with (say) having only five or even four London-bound trains in the 2pm-3pm slot? I don't see how you could argue otherwise. And that's consistently my experience across the old SR - off-peak services are well-loaded, especially given that so many operate in 8-carriage formations through the day, but never so much that going from a 10-minute frequency to a 15-minute frequency would hurt.

But I'm not convinced that model can be transposed to the GWR network. GWR doesn't operate many 6tph routes. Most major flows are 2tph at best. (Reading-Paddington is a special case, of course.) Putting a 'firebreak' on Oxford-London fasts, for example, might mean halving the mid-afternoon service. Anyone who's visited Oxford station at such a time can envisage exactly how well that might go.

We already have experience of this on the Cotswold Line. Even post-redoubling, there are major service gaps, gaps which make the train a much less feasible option for many journeys. 2pm-4pm on Saturday up from Charlbury is my own cause celebre: a 3pm service would make me much more likely to take an afternoon out in Oxford. The evening return has another two-hour gap, between 9pm and 11pm. And I realise that here in the Cotswolds we are blessed with an infinitely better service than the TransWilts or much of the rest of the GWR network.

So, firebreaks on Southern? Absolutely - if it's operationally worthwhile, go for it. On GWR? I remain to be convinced.

It needs to depend on the level of service. 1tph to 1tp2h no way, 6tph to 4tph would be ok

But he didn't say that!

So it's only short-term, and missing a service is only needed if there isn't spare capacity otherwise.

Hmm. Yes, he did, and more.

"At least a 30min break in its operation" equates to losing one train in a 2tph service, for one hour. In addition, that means losing 2tph in a 4tph operation, and 3tph (consecutive) in a 6tph operation

I think you're misreading that too. The "it" that has a 30 minute break in its operation is the train, as a physical object, not the service. Provided there are more trains is use that day than the service requires, and crew (e.g. at a kind of shift change, at least for that train), your 2 tph service can be continuous. He explicitly warns against running the maximum peak service all day, partly for than reason.

That's how (and why) I voted. If passenger usage thins out enough on a given route during the off-peak period, it makes sense to reduce the frequency a little from say a 6tph service (every 10 minutes) to 4tph (every 15mins) or 3tph (every 20mins) provided that this does not put off many passengers and would not result in passengers not getting a seat who otherwise would have one. However, any route which has less than 3tph should not be subject to any reduction in services, since intervals any longer than 30mins are likely to introduce many cases of "there isn't a train when I need it".

French Railways has, or at least used to have, a 'white period' of an hour or so in the middle of the day when no trains were timetabled to allow for inspection of the infrastructure. This was mostly on the trunk routes and the train service was never so intensive anyway - suburban services (at least on the Île de France) seemed to keep running.

Note, though, that the Gibb report only suggests this a short term measure until the other recommendations bring longer term improvements in reliability.

I votod no and would be dead against this.  Sure match the supply of trains to demand, but leaving a white space gap in the timetable just for service recovery is an admission that "they" are incapable of running a proper service.  Operating staff will know they have this recovery time and it will cause them to relax their focus on minimising delays.  Just as timetable padding encourages sloppy platform work 

As is well known, many rush hour services are overcrowded, some grossly so.
Complaints about such overcrowding are often answered with statements that it is the passengers fault to choosing to go to work in normal working hours.

Rather than provide sufficient capacity, it is often suggested that the answer is to encourage flexible working hours so as to spread the load.
A partial closure or significant thinning out of services during the middle of the day is hardly going to encourage off peak travel.
I have no doubt that some TOCs will simply use this as an excuse to cut back on staff during the fire break, and employ only the absolute minimum to just about run the much reduced service, with no staff in reserve to recover from any earlier disruption, when of course the whole point of the exercise was to facilitate recovery from earlier problems.

That might make sense on the London commuter network.  But away from the capital things are different.  Firstly providing the extra capacity ought not to be prohibitively expensive because we are talking about things like allocating a single extra unit to the Melksham line or running the Cardiff-Portsmouth service as 4 cars instead of 3.  Small relatively low cost changes which would make a real difference.  We are not talking about spending 15 billion on tunnelling under the centre of a city which is what providing extra capacity in the capital means.  Secondly, the regional railway is certainly not deserted outside of the peaks.  They are busy at commuter times but their passengers are not overwhelmingly commuters.  Off peak trains may not be full and standing but they are usually pretty well used with business, education and leisure trips, foreign tourists etc,  so those extra coaches are not often carting around fresh air.