Save the Melksham Train
This is an archived page.
Please see [here] if you're looking for the current "Save the Train" home page.
Please see [here] for the TransWilts site.
Friday, 26th March 2010. Launch of TransWilts Community Rail Partnership. Bridge House, Trowbridge, from 19:30
An exciting new step forward to rejeuvenation of the line, its service and its use
BUSY, EXCELLENT MEETING ... see [here] for initial report
Train service remains dire - southbound from Swindon at 06:15 and 18:45, northbound from Westbury at 07:02 and 19:35. Please pledge your support if you would like to see an increase to six trains a day -
arriving Swindon at 07:48 08:53 11:50 14:50 17:36 and 20:19,
returning at 06:18 09:02 12:02 15:02 17:55 and 18:45.

Keynote for 2010

Forum archive ... Blog archive

Train Service, Swindon to Westbury via Melksham
Open Access Proposal




In December, First Great Western withdraw the Swindon to Southampton passenger train service. The southern section of the line from Southampton to Salisbury, Westbury and Trowbridge will continue to be served by their other services (though with much more limited service to certain stations), but the northern section will be slashed to just two trains a day, in the very early morning and early evening.

The previous operator (Wessex Trains, part of National Express) who ran the service until 31st March 2006 grew passenger use 8 fold over 5 years - ticket sales at Melksham which is served ONLY by this line grew from 3,000 to 27,000 according to the office of the rail regulator, and annual journeys on the service now number some 109,000 (figures from First) - an average of 32 passengers per train.

With First giving up the service almost entirely, the way is left open for another operator to provide service on this fast-growing route under open access. The towns of Swindon, Chippenham, Melksham, Trowbridge and Westbury are all experiencing rapid growth in both housing and business use, and the "West Wilts Corridor" is an important and logical one, a part of the county council's key growth strategy. The main road servicing the corridor - the A350 - has been improved over the years but is severly congested. The road around Chippenham is solid in the rush hour, and sections through Beanacre and past the North of Melksham are busy throughout the day. Safeguards over land for a proposed bypass for this Melksham / Beanacre section were withdrawn from the next 10 year plan.

With a service of just 5 trains each way daily at present, the average Melksham inhabitant makes just one passenger journey from his/her local station for every 20 journeys made from the station in the nearby town of Bradford-on-Avon, which enjoys a service roughly once an hour in each direction.

The current train service has grown dramatically, in spite of a fearsome reputation for frequent cancellation (and that continues under First, with 4% of services not running in the last 3 weeks prior to this report being written), and replacement of trains by buses for engineering works (at least some trains have been replaced buses on the majority of weekends over the past year). Publicity for the service has been almost non-existent, even at Swindon and Chippenham stations where it was a "guest" service, and connections have been problematic when main line trains run late.

Why has it been so successful in spite of the dreadful service? Alternative "end to end" services for the line involve a dogleg journey via Bath, with an awkward change and a very large number of steps. Some recommendations from national web sites even route passengers from Salisbury to Swindon via Basingstoke and Reading. Interchange from train to Bus at Chippenham adds 50 minutes to the journey time, with the bus not serving the station and a long walk or extra connection involved. And many of the passengers using the service are "core" users who do not have access to their own transport and who can't be easily displaced. The "choice" and "leisure" markets remain largely untapped.

There is no need for the service to be cut. There's a golden opportunity for an experienced rail operator to run services under an Open Access arrangement. This document goes on to provide further detail.


  15 minutes
  Also served by London - Bristol Expresses
  10 minutes
  No other passenger service on this section
  10 minutes
  No other passenger service on this section
  6 minutes
  Served by Bristol to Westbury trains (many extended)
  Current service from Frome (a.m.) and some to Salisbury and Southampton

  Connections to Gloucester and Cheltenham, South Wales, Reading and London.
  Very large industrial, commercial, shopping centre.
  Rapid growth
  Parking and congestion issues; station well located in centre.
  Additional main line platform added in recent years.

Double track to ... Wootton Bassett Junction (where South Wales line leaves)

  No Station at present.
  Suggested for Park and Ride.

Double track to ... Chippenham

  Connections to Bath and Bristol.
  Industrial and commercial town.
  Rapid growth
  Station well located for town.
  Two operational platform faces; trains can be turned (and quickly) but not clear of lines
  Additional platform face available, but without track.

Double track to ... Thingley Junction

Single track to ... LACOCK

  No station at present.
  Suggested for tourist station for National Trust village and local catchment.
  Station would not be central to the village,

Single track to ,,, MELKSHAM

  Unstaffed station.
  Residential and commercial town with some industry.
  Rapid growth
  Station situated about 1/2 mile from town centre
  Many suggestions for station improvements
  Short platform / no Grandfather rights, so long trains require selective doors.

Single track to ... HOLT / STAVERTON

  No station at present.
  Suggested for station for North Trowbridge and for Holt.
  Tourist potential for very popular National Trust property "The Courts".

Single track to ... Bradford South Junction

Double track to TROWBRIDGE

  Station, two platform faces.
  Connections to Bath and Bristol.
  Residential and commercial town with some industry.
  Also County town.
  Local paper describes Trowbridge as "Boomtown".
  Reasonable situation for town.

  Large commercial development area.
  Suggested for station for access to businesses in tha area.

Double track to WESTBURY


  Station, three platform faces in use
  Connections to London, Taunton Exeter and the South West
  Also connections to Salisbury, Southampton and Portsmouth
  Also connections to Frome, Yeovil and Weymouth.
  Station is away from town centre. Growing town, major traffic issues.

Onwards from Westbury, some current service run via DILTON MARSH and WARMINSTER to Salisbury, Romsey and Southampton.


We calculated in January that the current service makes neither a significant loss nor a significant profit, based on evidence available at that time. We looked at three options for future service between Swindon and Westbury:
  a) A continuation at existing levels / timings (approx)
  b) A regular service of 8 trains per day - 2 hourly for the most part
  c) The service that First will be running from December - 2 trains per day.

Option c

With just two trains a day, current passenger usage will plummet.

There is a peak commuter flow into Swindon in the morning and back out in the late afternoon, but the passenger already complain that the morning train is too early into Swindon at 08:18 - they preferred previous timetables when it arrived there just after 08:30. From December, we understand the arrival will be well before 08:00. In the evening, the current departure of 17:45 which suites most current users will be put back by an hour. We have not identified any significant new traffic flows that the new service (nor its counterpart trains from Swindon at around 06:20 and arriving back there at around 20:15) will service, though they may cater for a potential London commuter traffic in the future.

Other flows on the line are many and varied, with regular users on all trains that will be withdrawn. A proportion of these other flows use the existing "commuter train" that will continue to run at one end of the day, but other services for the return journey. It is unlikely that a significant number of these travellers will wish to use the train in one direction and make alternative arrangements for the other direction.

Net result - a "ghost" service with very few passengers. Low expenditure compared to other options, but a minimal income leading to a significant loss.

Option (c) also provides minimal community service value and would create some significant hardship.

Option a

With the current service level and regime continued, we forecast that growth will continue, perhaps a little below current levels. Train capacity remains available and sufficient for passengers for the next 3 to 4 years, after which some train strengthening would be required.

With the additional income, but with little extra expenditure, a progressively better financial return will be made. With double the current ticket sales in 4 years time (so twice the income), "break even" turns to a healthy profit.

There is, however, a problem with option (a). The stock operation is linked in to onward services to and from Southampton, and other routes, and with the major revisions elsewhere there will not be a train available at Westbury when needed to run it. As there is never more than 1 train on the line at a time, a dedicated train COULD be provided, but it would sit idle for periods and would require awkward shunting to clear platforms at Westbury at times.

Option b

Trains from Westbury at 05:45, every 2 hours to 11:45, 14:45, every 2 hours to 20:45.
Return from Swindon at 06:45, every 2 hours to 12:45, 15:45, every 2 hours to 21:45.

Talk to passengers and they'll tell you that they want a regular, reliable service. Talk to POTENTIAL passengers and they'll tell you they would use the train if they knew about it / knew when it ran / could rely on it.

With a dedicated train, reliability would be much improved; incidents from Portsmouth and Brighton to Cardiff would no longer have the same knock-on effect. With a regular (clockface) timetable, the less frequent traveller would be able to plan his journey more easily and would be able to use the train without having to enquire every time.

The service suggested in this option meets the aspirations of current passengers with an appropriate alternative service for virtually every current train. If it continued to run, The commuter train at into Swindon at around 07:45 and returning around 18:45 would add a welcome peak time strength, provided that interchangable ticketing was available, and minor timetable adjustment might be needed to enable the two trains to pass at either Trowbridge of Chippenham.

With reasonable marketing, this service would grow in excess of historic growth levels (and with good marketing, it would boom!) and rapidly turn in a healthy profit. The rest of this document goes on to consider other aspects of option (b) - the preferred option.


With the addition of more platform capacity at Swindon in recent years, and platforms at Westbury already signalled for multiple trains in a platform, there should be no problem in turning the train at either station. No other passenger trains use the single track section of line from Bradford South Junction to Thingley junction, and there is adequate capacity on the sections from Westbury to Bradford South, and Thingley to Swindon.

The single track line is shown in Network Rail plans as a medium growth line. From just an occasional freight train a few years back, this traffic has already expanded. Should it continue to grow, additional signalling may be required so that trains can follow each other without the need for passing loops or doubling. By using existing layby facilities at Westbury and Swindon, longer distance freight trains can follow passenger trains through - with a one-hour northbound "slot" followed by a one hour southbound slot, a 20 minute journey time and a 10 minute interval between trains, maximum capacity in a 2 hour period would be 4 trains in each direction. 30 trains per day in each direction is comfortable under this regime. Note that the 20 minute journey time would allow additional passenger trains stops to be added at Lacock and Staverton/Holt.


There is minimal train servicing at either Swindon or Westbury.


a) Stagecoach's Salisbury depot is 30 minutes by rail south of Westbury on a double track line and would be the logical home depot for the service. Note - the SW trains franchise is up for renewal, and future operator co-operation would be needed.

b) With the co-operation of First, a train could be 'fed' onto the line in the early morning attached to the Stroud Valley service and be 'collected' at the end of the day. This might involve the loss of the 05:45 from Westbury, but gain an extra 22:45 from Westbury.

c) First's new Bristol Depot, Cardiff (Canton), Oxford.


The current timetable shows 4 trains each way on Saturday and 3 on Sunday, but they rarely all run. Weekend trains are busy, (in some cases busier than the weekday equivalent), but when a substitute bus is run, it gets very few passengers - perhaps a quarter of the number that use the train. Part of the problem has been extended Network Rail maintainance programs in recent years, and part has been the replacement by bus of the entire section from Swindon to Westbury, even if it would have been possible to run Westbury to Chippenham or Swindon to Melksham trains. The route is also used for diverted trains from other routes.

We propose a similar service on Saturday and Sunday to during the week, but without the round trip from Westbury at 05:45 on Saturday and Sunday. Sunday is becoming far less of a rest day, and the trains will be busy with weekend shoppers going in to Swindon, staff for those shops, and for a wide variety of other weekend travel. Weekend trains will bring a tourist traffic too .. and much improve that traffic is and when Lacock and Holt/Staverton stations re-open.


Melksham station is currently poorly situated in the rear of an industrial estate with no other public transport services running there, with no telephone, no taxi rank, and with poor signage. Th helpful map to the town centre has recently been replaced by a photograph of the local train manager. There are about 10 parking spaces, for which no charge is made.

The Melksham Railway Development Group tend flowers and keep the station tidy.

Land has been set aside for a new station 400 yards to the North of the current site, accessed from a new area of housing and very close by to a restaurant that's open from 7 a.m. to 11 p.m. 7 days a week, and a large retail store. There is also land current in the hands of BRBR (but due for sale by auction on 31st October) for access to the new site and along the line to the current site.

Options / suggestions

i) Utilising the current site, route the town bus (sponsored by the local town council, very smart, operated by a local company) to the station each hour to connect in and out of the train - ideally call 10 minutes before the train leaves to Swindon, and 10 minutes after the train arrives from Swindon.

ii) Complete a link road to the new area, which would allow the local bus (and service 234, run by 2 different operators, one of whom is also the current train operator) to pass by as part of a loop.

iii) Provide a bus stop on routes X71 / X72 / 272 (in effect hourly services) that
pass by the station at the head of the approach road. At present, these buses don't
stop for about 400 metres (out of town) and 800 metres (inbound) and the nearest stops do not indicate the proximity of the station.

See "The line in question" above for other stations.


a) If operated from Salisbury, a two-hourly service with one appropriate / fast train might be able to serve Salisbury, Warminster and Dilton Marsh before Westbury. Current timings don't make this appear practical, but I'm told that the current schedule is very leisurly so an extension from Westbury to Dilton Marsh and Warminster (turning issues on busy line? no bay platform) and perhaps Salisbury should be considered.

b) Extending from (a), the service might run into the Southampton - Salisbury shuttle and / or the slow Waterloo services that currently terminate at Salisbury.

c) With a train calling at Melksham every hour, an hourly connection (via Trowbridge or Chippenham) is possible to Bath and Bristol - the destination of many people. With appropriate publicity / timetables, much business to be had.


Salisbury, Southampton, Yeovil, Oxford, London via Salisbury
Wootton Bassett, Lacock, Staverton/ Holt, White Horse, Radstock

Each of these are significant opportunities, though perhaps not in the initial service.

An Oxford extension from Swindon would revive a much-used service that was withdrawn as a Bristol - Oxford train due to capacity issues in the Bristol area, and with a Salisbury extension would provide a natural route for tourist as well as regular traffic.

Radstock is an important commuter feeder to Bath.

We will be happy to expand on this section for individual organisations upon request. There are many possibilities - see (for Example) the May 2000 Parkman report.

Currently, local fares are around 18p per mile (compared to much higher rates on main line services from other local towns - around 50p per mile for open standard class returns). Long distance fares reflect the main line prices.

Local fares could be increase by around 25% in real terms and little traffic would be lost. Introduction of a higher fare at the time of a better service being started would not be an issue for most users, though this document costs a more gentle increase.

There's a significant opportunity for travel from Swindon, Chippenham, Melksham, Trowbridge and Westbury to London Waterloo via Salisbury, if that service is provided at a non-premium price. The current open return fare from Chippenham to London Paddington is 97 pounds and if a fare from Chippenham to Waterloo could be offered at 60 pounds, even with the longer journey time there would be significant uptake. (Notes - 48 pounds day return SALISBURY to Waterloo, 9.40 return Melksham to Salisbury, and for travellers headed for many parts of London, Waterloo is much more convenient than Paddington.)


* Year 1.

2 x 153 Unit hire and service 300,000 pounds
Crewing of train 287,000 pounds
Line access charges 100,000 pounds
Publicity, admin, etc - 50,000
Total expenditure - 737,000 pounds

120,000 journeys at 5 pounds each = 600,000 pounds
Total income - 600,000 pounds

* Growth assumptions.

a) Operating costs in line with 5% inflation
b) Traffic growth in line with historic 35%
c) Ticket prices growing at 7% (2% above inflation)

* Year 2

Expenditure - 774,350
Income - 162,000 journeys at 5.35 - 866,700

* Year 3

Expenditure - 813,068
Income - 218,700 journeys at 5.73 - 1,253,151

* Year 4

Expenditure - 853,721
Income - 295,245 journeys at 6.13 - 1,809,852


a) Current levels are already over 100,000 passengers per year so that first (start up) year shows a lower growth than historic and subsequent years - a "blip" for the changeover

b) Historic Passenger growth has been acheived on poor reliability and marketing and an irregular service pattern that people cannot remember. The service only attracts 1/20 of the journeys per head of population from Melksham than the next line across does from Bradford-on-Avon. There is plenty of scope for the growth forecast above and indeed even higher growth should be achieved if the historic issues are dealt with. I do NOT see 4 years ahead as being the end of the growth curve.

c) The figures quoted are based on information supplied by First and Wessex Trains and checked with other sources where applicable. They are the incremental cost of providing this service, and its true income. They do NOT include additional administration resources, nor the costs of working train from a remote depot to the line, nor issues with a stand in train. If the service was not to be run by the current operator, the "best value" here might be a train operating out of South West Train's Salisbury depot. For consistency of fleet there, a more expensive 2 car unit might be provided.

d) I understand that Wessex trains may only have received 40% of the income from the service when they ran it, under an arrangement known as ORCAT. However, they also got a proportion of income from Chippenham - Swindon from other operator's trains. I am NOT able to factor such issues into the figures above, nor to comment on through ticketing, etc.

e) This service is NOT price sensitive within a few percent; I've allowed 2% above inflation for 4 years, but I believe that 3% or 4% above inflation would not lead to any significant business loss.

f) Average train loads (currently 32 per train) rise to 52 per train by the end of the period above. At that level, overcrowding would become an issue on busy trains with a 153 unit, and a switch to a 158 or 170 would be appropriate. Comfort WILL be a factor in attracting passengers on the faster sections, and I would be reluctant to suggest switching to some of the cheaper alternative 2 coach trains. Increasing the service is also an option but would involve extra infrastructure work to increase line capacity.

g) Service could usefully be extended to Salisbury (Southward). I have NOT studied the full financial case but this is a section that's also being trimmed back heavily by FGW, against reports commissioned by WCC that recommend it be stepped up to half hourly. There is significant traffic from West Wilts to and from Salisbury and the extra service would not only bring in revenue from the extra section, but through traffic that would not otherwise be on the line too. A second train set would be required for this service to run two hourly. Other options to increase income include re-opening stations at Holt and Lacock (heavy potential tourist traffic, especially if service runs from Salisbury). These are attractive as a 2-hourly service WOULD pull passengers in, and the new stations would just be single track structures.

h) The figures assume that the 8 trains per day are the ONLY passenger trains operating between Trowbridge and Chippenham. Competition from other services (say from Swindon at 06:20 and 18:30, returning from Westbury at 07:00 and 19:30) would if anything enhance both offerings by giving more choice to users before the peak hour (a.m.) and after (p.m.), only concern being that this service would break the "clock face" timetable proposed.

i) Based on figures / report prepared by Graham Ellis, January 2006. This update, 21st July 2006. Although the initial report made some educated guesses, information provided subsequently has shown it to be remarkably accurate.

j) Report updated 28th July to add in the hire of a second class 153 train to cover time that the first unit is unavailable for service, and to update the crewing costs slightly. Although the spare unit doubles the hire cost, other costs are unchanged or only marginally changed. This change makes for a much more reliable service which can only aid the growth under a new operator. Strong marketing of this spare resource will give users who have tried and abandoned the train in the past to return all the quicker. The "spare" unit could also be used as backup for other services if based at (say) Canton or Salisbury from where other similar units may be operating, and thus the cost may be spread.


The costings just provided do not include any capital investment and pump-priming funding, and there is a case for working with the SWRDA, the County Council and other bodies on such. This also relates to some of the developments discussed for Melksham station.

If the line needed to be doubled in the future, most of the trackbed from Thingly to Bradford remains in situ; price for a "lowcost" option of 26 million pounds has been quoted. "It should be remembered though that , in some industry quarters , there is a feeling that Network Rail over - estimate the cost of infrastructure projects."


For further background information on the statistics used to compile this report, please see and / or use the contact details. Figures quoted / used are from reliable industry sources and figures that have a doubtful basis (such as to small a sample) have been avoided. Estimates are on the conservative side.


The "Save the Train" campaign that's advocating this Open Access Proposal has support from local politicians of all three major parties, from councils at parish, district and count council level, from the SWRA, and from the users of the current service.

All 4 MPs who represent constituencies along the Swindon to Westbury section (both Labour and Conservative) spoke in Westminster in April against the planned cuts, and we've also been offered the support of the Lib Dem member for Romsey (covering 2 stations to the South on the current service) amongst others. Senior opposition politicians have also offered their active support.

Senior industry insiders have looked at and studied the figures on which this proposal is based, and say that there's "nothing wrong" with them. Operational staff on the service are also very helpful and supportive.

We've signed up several hundred users and interested parties via our web site, which has been visited by some 9,000 separate host computers since it was set up in the latter half of last year.

Many thank too to everyone who's provided information that's helped in the preparation of this document.


There is a strong case for a regular train service provided by an open access operator on the Swindon to Westbury route, from December 2006.

The route provides a much needed service, has extensive local support, and is already a success story. With First moving out of the market, there's a gap for another operator to move in to a potentially lucrative market.

Widespread support will be given to any operator who provides a regular, reliable service at a realistic cost.


For further details, please contact Graham Ellis -

This site is hosted by Well House Consultants Ltd. (
Contact Information

Further Information:
 Current Summary
 Daily update
 User forum
 Service now
 Service future
 Future Analysis
 Recent Statistics
 Recent letters
 Letter to DfT
 Save the train
 Support us
 Other Maps
 Station facilities
 Station approach
 Trains diverted
 About Melksham
 Site Map
 About this site