What is a reasonable connection?

[whole diary]

[ChrisB]

What I meant is that I doubt a train to every destination from Reading can be timetabled within 15 minutes of arrival from every train. Thus achieving a 15 minute connection is impossible from anywhere to anywhere else

Point taken where you are connecting between a service or services with hourly (or more than hourly) frequency of stopping at Reading. But many services are (or will soon again be) more than hourly, and as who has spent 20 odd years changing between two of these (Thames Valley from Goring to Reading, North Downs on to Guildford and back in reverse)

So those from The Berks & Hants might have to wait longer to suit the Thames Valley (for example) Or from the Cotswold Line, for example? Why should your connection be more suitable?

What I'm saying is that so many connections in so many directions are made at Reading that to make everyone happy with less than a 30-ish minute connection is just impossible....

[Marlburian]

Further to my post of yesterday, time was when I took a train from Tilehurst to Twyford or Maidenhead and did a circular (rectangular, actually) walk back to Twyford or trekked backed through Marlow to Henley and then took a train home. On average I had a delay of 15 minutes at Reading for a half-hourly Didcot stopper, but with the current hourly service I fear that longer than that would see the joys even of Reading Station lessen considerably.

Faced with a wait longer than 30 minutes or so, I did take the bus occasionally, but I don't fancy that at the moment.

[stuving]

All this talk about connections vs direct trains reminds me of Network Rail's much-maligned "Improving Connectivity" idea. That report has migrated from the internet into its locked cellar, so I can't provide a link to it. But, apart from being discussed on this forum, this page from RailFuture is still there, and says this:

Thirdly, whatever its merits it seemed to many that IC^▸ (Inter City) should have adopted a 4th and overriding principle. Having identified ‘major hubs’ in the East Anglia network it should have developed a service pattern between them that did not then require a change of trains en route. The prime example was Ipswich to Peterborough (for ECML^▸ (East Coast Main Line)) which relied on changing at Newmarket, an inconvenience bound to surrender even greater market share to the A14! It also overlooked the potential for additional platforms at Ipswich in favour of connecting at Stowmarket or Manningtree. Operational convenience seemed to carry more weight than passenger experience (or capital cost!).

Fourthly, its obsession with the avoidance of duplication read like a plan for cuts, “How do we do the same with less or more with the same”. It did not sit comfortably with stakeholder ambition for improved services and reduced overcrowding in a fast growing region. Poor though many local connections are, East Anglia’s biggest ‘connectivity’ problem is the lack of through services to other regions unless via London.

As ever, you need to read that original report to understand why they proposed that, and that removing duplication per se wasn't an objective.

On the other hand ... this is from SNC Lavalin, who appear to be touting for business as timetabling experts:

Without intervention, the rail market will continue to lose out. Too many people are choosing other modes of transport because end to end rail journeys are ill-coordinated and don’t meet customers’ needs. A long wait for a connection can be the tipping point when deciding what mode of transport to take. The detail of the timetable can make the difference between a journey taking three hours or five hours and resulting in the car being the quicker and easier option. Given that only 1% of possible rail journeys on the network are currently served by direct trains, the importance of good interchange cannot be overstated.

A connection-based timetable is entirely feasible in Britain. It has a proven track record in both Switzerland and Austria, whose rail networks have some of the highest market shares in Europe.  The key to unlocking Britain’s potential lies in designing the timetable earlier in the infrastructure development process.

Our existing process, where infrastructure plans are passed from Network Rail to government for funding without designing the timetable that will eventually operate on them is a barrier to this change. These plans state the expected future service levels (i.e. how many trains per hour will operate) but not the timetable and, therefore, are incapable of considering the needs of people who interchange.

I'm not convinced by what they say about how infrastructure is planned now. But given the vicious circle of missed connections making people value direct trains highly, and timetabling around direct trains making the connections bad even when not missed, there is something in their approach.

[stuving]

Does anyone have any record of the outcome of that "Improving Connectivity" consultation? I've got the proposal report, but nothing from after it - and neither did it get onto our thread. But it's not dead, at least not so far as its sponsors are concerned.

The work at NR^» (Network Rail - home page) that led up to it was supported and partly staffed by Atkins, under the title "Retiming Britain's Railways". One of the people involved, Chris Nuttall, who is now at Atkins, is still banging its drum (or was a year ago). Here's a bit of a piece of his in Global Railway Review:

Retiming Britain’s Railways (RBR) is a unique Atkins-sponsored proposal to revolutionise rail travel by slashing journey times, dramatically improving frequencies and relieving overcrowding. The crux of RBR is the introduction of a new, carefully planned, connections-based timetable designed to eliminate unnecessary time waiting at interchanges and transform a multitude of journeys across the country. Its goal is to enable people to travel seamlessly between any two stations on the network, irrespective of whether an interchange is involved.
...
The public are clearly on board with the concept. In 2014, Network Rail published ‘Improving Connectivity’, a public consultation which focussed on RBR’s Anglia proposal and received an overwhelmingly positive response.

Awfully hyperbolical, isn't it? But I wonder if he is a one- (or few-)man band, or whether there is a kernel of people in NR advocating this approach. For example, have DfT^» (Department for Transport - about) been seduced into thinkiing this can help them find some small change to placate the Treasury, while adopting "retiming" as a mantra can sort out (for example) Salisbury. Even if there is a year to wait before any retiming ...

Those in fact are the same words as SNC Lavalin's, which were undated.

[Rhydgaled]

I voted 'other': it depends!

For me a big consideration would be the environment I had to wait in - a station with decent roomy waiting room, toilets and refreshment facilities and maybe some interesting architecture, or a bleak windswept platform with a bus shelter/unofficial urinal.  I'd also be less worried if I was changing onto a train starting at the interchange if I could get on at leisure well before the departure time.

and that is one of the reasons it depends. A 'connection' at a bleak windwept station is only reasonable if services run every 10 minutes so you never have to wait longer than that. With a decent waiting room and toilets, both of which are actually available for use and not closed or behind a barrier, then 20-30 minutes is ok.

[Red Squirrel]

I should probably have voted the same way as Rhydgaled, but I plumped for 15 minutes.

I access the national rail network via the Severn Beach line, so I always have to make a connection. Under the current timetable, connections to and from this line vary from good (i.e. sub-15 minutes) to terrible. As an example, Western Pathfinder and I were faced with an 85 minute connection on our mid-evening return from the Coffee Shop AGM^▸ (Annual General Meeting) last week. This seems pretty poor for an urban metro service. The previous week my family had a 55 minute connection when returning from Bath.

Perhaps what's 'reasonable' depends on the length of the journey. If I was travelling to Nottingham, a 15 to 30 minute change at Cheltenham would probably not seem too bad. On a recent trip from Bristol to Chepstow though we spent considerably more time waiting for trains than riding on them, which didn't seem at all reasonable!

[grahame]

I will be following up on the results of this poll in the next couple of days.

In the meantime:

From https://www.railforums.co.uk/threads/swr-cancels-saturday-salisbury-waterloo-bristol-service.223153/page-3#post-5371679

The morning SWT^▸ (South West Trains) train to Waterloo had a journey time from TM^‡(resolve) of exactly 3 hours, 08:49d 11:49a

The nearest equivalent via GWR^▸ (Great Western Railway)/SWT will take 2 hours 56 min 09:24d 12:20a

So despite having to change at Salisbury and Basingstoke, it’s actually slightly faster.

Do members feel this is a reasonable alternative for passengers from north west of Salisbury to London?

[ChrisB]

I would still be using that alternative, yes