How to provide capacity and increase service to hourly each way via Melksham

[whole diary]

[grahame]

The problem - an all stations passenger train every 2 hours each is too infrequent for most potential users. Although journeys rose from 3,000 to 75,000 per annum when the service increases form 2 to 8 then 9 trains each way per day, comparison with other similar local town suggests that with an hourly train passenger journeys would increase to over a quarter of a million journeys a year. That conclusion is backed up by the community / survey feedback which indicates that there are far more people who say the service is too poor for them to use than actually use it.

However, the line is a single track without and passing place or even the ability for a second train to follow the first train until the first has cleared at the other end.  It's at close to full capacity, especially when you consider the number of freight trains that also use it, long distance expresses that use the line from time to time, and the likelihood of those extra trains turning up late, early, or not scheduled on the line but needing to use it.

Provision of a loop to let trains pass each other at Melksham would be expensive due to the cost of a complete second platform and on an access for all footbridge.  Provision of a short loop just away from the platform at Melksham would slow down local trains as they would have an extra stop to make.  Provision of a long ("dynamic") loop is a potential solution.  Adding an extra signal on the route to allow two trains to follow each other would lead to complex signalling and problems when trains arrived out of sequence.  Terminating two trains end to end at Melksham and having all passengers transfer both ways has been done at Ormskirk but takes time and puts people off making the through journey which accounts for two thirds of the traffic.

 SUGGESTION - A loop at Melksham to allow passenger trains to pass each other. Single long platform.
 Example similar to Penryn (Cornwall), Dovey Junction (Wales) and Bad Doberan (Germany)

 Already planned service from June 2024 (Monday to Friday) - time (midline) of Melksham departures
 To Swindon   07:21   08:02   10:02   12:32   14:33   16:38   18:53   20:21   21:31
 To Westbury   06:32   09:10   11:31   13:40   15:40   18:00   19:09   20:40   22:56

 1. Train from Swindon arrives at Melksham to the north of the current platform
 2. Train from Westbury arrives at the current platform
 3. Train to Westbury leaves around a loop past the Swindon train and current platform
 4. Train to Swindon leaves, passing the north plaftorm
 
 The suggestion would:
 ADD^▸ (Automatic Dropping Device) to Swindon   06:34   09:12   11:33   13:43   15:42   17:32
 ADD to Westbury   08:00   10:00   12:30   14:30   16:35   18:30
 
 So Proposed Monday to Friday - Total service to Swindon:
 06:34   07:21   08:02   09:12   10:32   11:32   12:32   13:33   14:32   15:42   16:32   17:32   18:53   20:21   21:31
 Proposed to Westbury
 06:32   08:00   09:10   10:30   11:30   12:30   13:30   14:30   15:40   16:30   18:00   18:30   19:09   20:40   22:56
 
 Proposed Weekend - Total service to Swindon based on rest of the week service:
 06:34   08:02   09:12   10:32   11:32   12:32   13:33   14:32   15:32   17:32   19:30   21:30
 to Westbury
 08:00   09:10   10:30   11:30   12:30   13:30   14:30   15:30   16:30   18:30   20:30   22:56

How would it work? - more detail

* With a couple of exceptions, trains would be clockface, leaving Melksham in both directions within minutes of each other every hour. 

* Service on BOTH main lines at Trowbridge and Chippenham are already clockface so the extra trains would slot into paths between other trains already there.

* Westbury terminators in the extra hour would occupy the platform used for turn around in the other hour, so would be within capacity.

* Westbury arrivals are timed to be ahead of the service to Portsmouth and to leave after the service from Portsmouth, giving an hourly connection to and from Warminster, Salisbury, Southampton and beyond.

 * There is only a couple of minutes of extra time in the hour occupied by the local train service.  With the loop long enough that gives a doubling of capacity for other trains such as freight and passengers on divert

* Multiple trains could follow in the same direction giving a 10 rather than 18 miute headway if nothing is coming the other way.

* Services up from 9 to 15 each way Monday to Friday and to 12 each way on Saturday and Sunday.

* Existing Peak and education trains remain pretty much unchanged; these are highly tuned already.

* Three minute wait for operations and robusness in the event of minor delays at Melksham.

* New services fill peak gaps - early to Swindon, traditional peak services to and from Trowbridge and Westbury.

* Same service base 7 days a week - the main lines are, so surlely in should be?  09:10 soutboud service currently extends on Summer Saturdays to Weymouth and perhaps could do so far more (all year?  Sundays?  Even during the week?). Likewise one of the later returns.

* Services would co-ordinate with electric town bus "15" leaving the station at xx:35 (and with authority to wait 5 minutes for train) and returning there at xx:20, giving more than adequate 45 minutes for the journey to the Town Centre, Sandridge Road, East Melksham, Pathfinder Way, Bowerhill Industry and Hampton Park West.


Comparative data for usage of train service (2018-19 to 2022-23), calls per day, town population
Grand total - journeys to and from Chippenham   1972350   1937240   373498   1124226   1444888   (94 - 45000)
Grand total - journeys to and from Trowbridge   933894   917458   252180   640376   734772   (111 - 37000)
Grand total - journeys to and from Westbury   548720   557764   152218   438970   518998   (152 - 16000)
Grand total - journeys to and from Bradford-OA   534086   550720   117532   351558   426702   (93 - 9500)
Grand total - journeys to and from Warminster   348658   355386   103082   283906   295484   (73 - 18000)
Grand total - journeys to and from Frome   201332   207704   47934   166496   207090   (31 - 28000)
Grand total - journeys to and from Melksham   74534   75292   18806   52342   64424   (18 - 26000)



Picture - train from Wismar to Rostock (in the foreground) pulling out to pass the train from Rostock to Wismar that sits up the platform and will leave once the train going the other way to Rostock has passed it.

Even the curve of the track is right, and there is space where there used to be a second track at Melksham!

Picture would read "train from Swindon to Westbury (in the foreground) pulling out to pass the train from Westbury to Swindon that sits up the platform and will leave once the train going the other way to Westbury has passed it".

[grahame]

Addendum

1. As a result of interaction on social media to help show how this could work, here are a photograph of current Melksham Station and a sketch showing where the extra platform and line would go





Green - existing platform (to become Swindon Platform)
Blue - extended platform (to become Westbury Platform)
Magenta - loop

2. A couple of timing adjustments to be made (no more that 10 minutes) to make evening trains the same time every day of the week. 20:30 weekend / 20:40 weekday to be a single time. 21:30 / 21:31 to be consistent weekdays and weekend

[IndustryInsider]

I hope they tamp that new track before it’s brought into service. 

[IndustryInsider]

I think freight might be your biggest friend in terms of this getting off the ground.  There is so much of it about now that it really is starting to affect passenger train performance.

I guess the trouble is that there’s a number of competing pinch points that should, arguably, be higher up the list…and money is tight to say the least!

Two examples that I would put as a higher priority are:

1) Didcot, where a complete remodelling east of the station with a grade separated junction is very much needed.

2) Leamington to Milverton (Coventry) doubling where the parallels with the TransWilts are clear.  Kenilworth being the intermediate station in this case, though that should be quite easy to add a second platform to as passive provision was made when it re-opened.  The biggest difference is the number of trains using it - usually 4 passenger trains and a couple of freights every hour, so it’s more of a key strategic route than the line through Melksham IMHO^▸ (in my humble opinion).

I think the best you can hope for in the shorter term is a couple of extra signal sections in each direction, so trains can be flighted through two or three at a time in one direction then the other.  That would definitely give you better capacity to deal with extra diverted trains as well as the local service.  Though whether it would be enough for an hourly TransWilts service would need careful thought.

[Oxonhutch]

1) Didcot, where a complete remodelling east of the station with a grade separated junction is very much needed.

I submitted the Didcot plans some years ago, but the authorities have still not considered it

[grahame]

I think freight might be your biggest friend in terms of this getting off the ground.  There is so much of it about now that it really is starting to affect passenger train performance.

Yes, I would agree with that.

Without freight, the line would probably not have survived at all - it would likely have been ripped up some 50 years ago. Without freight today, an hourly passenger service each way would be possible, but very inefficient indeed in needing multiple trains with longish layovers - but that is utterly hypothetical as the freight is a key part of the line and we would not wish it to go away in any form.

Extra intermediate signalling sounds attractive, but I worry about how robust that would be when trains don't run to timetable - especially freights.  I have take a look back at yesterday on Real Time Trains.  Yesterday was a good day - with no cancellations to the passenger services.    Twelve none-passenger trains passed through that did not call at Melksham, in addition to 2 expresses (the sleeper) and the eighteen passenger services - a total of 32 trains.  Only two of the none-passenger trains were more or less on time - eight were late enough to show up in red on Real Time Trains, and two were early enough to show up in cyan.   

I appreciate that quite a number of yesterday's trains were during the night, but I fear that adding many more (unless freight trains start running to time) and running them in flights, would be a reliability nightmare.   Even if schedules are such that trains have slack built in to their timetables to wait for the flight, there is nowhere at Chippenham or Trowbridge for them to wait off the main line.     I would be delighted if experts could re-assure me on this point; to my amateur eyes, extra track to allow for more trains to get through and trains to pass each other / wait on the current long single section sounds attractive, and as a previously double tracked broad gauge line there is potentially the space, though considering cutting and embankment decay, the Avon bridge and the profiled junctions full double tracking would be expensive.

I would agree that Kenilworth and Didcot also need attention - and there are others in the wider UK^▸ (United Kingdom) picture.  Peterborough has been done, I need updating on Ely and I have questions in my mind about future Birmingham to Manchester capacity when the high speed line from West London to Birmingham opens.  But several projects can be financed and undertaken in parallel, can't they?   Something of a rolling program but, please, not a single stranded one!

[Red Squirrel]

The Severn Beach line makes the point that signals are not a substitute for double track. It’s s fine when everything’s running on time, but when things get out of whack it’s a different story.

[Witham Bobby]

The Railways Inspectorate used to carry quite a lot of resistance to section-split signals on single lines.  A hangover from what happened at Foxcote on the Somerset and Dorset Ry in the 1870s