Resurrecting Railways: lessons from two Scottish projects

Peter Heubeck

The DfT’s “Restoring Your Railway” initiative has raised hopes of extending England’s rail passenger network. Yet Scotland has led the field in expanding its passenger network. This article describes the two largest completed projects and highlights best practice.

Part One – Airdrie to Bathgate

Growth and decline

South of the Edinburgh to Glasgow main line is an undulating plateau on which the line between Airdrie and Bathgate was closed in 1982 when trains to a car distribution depot became the only user of the Bathgate branch. However, services between Glasgow and Airdrie were transformed with electrification in 1960.

The ScotRail Whirlwind

In 1984 and 1985, Chris Green implemented a dynamic, business-led approach on Scotland’s railways. Passenger numbers grew in response to improved service quality, competitive pricing and a strong brand identity. With an enthused workforce, a supportive public, and a strong partner in Strathclyde PTE, ScotRail was growing.

Lothian Regional Council and others contributed to a low cost scheme for an hourly service between Bathgate and Edinburgh, calling at Haymarket, Uphall and Livingston North. Incredibly, the work cost only £1.3 million! The service began in March 1986 and was instantly popular which led to railway managers, politicians, civil servants and campaigners viewing rail re-openings more seriously. At the west end of the “missing” route, the line was extended to Drumgelloch, on the east side of Airdrie, in 1989.

The case for reopening and electrification

Local employment opportunities were limited. Acknowledging the need for improved transport links and economic regeneration, Government funded the Central Scotland Corridor Transport Studies which, in 2003, led to the Scottish Executive deciding to invest in a double-track, 80 mph, electrified through route from Edinburgh to Glasgow via Bathgate and Airdrie, with a 15-minute service frequency to offer hugely improved public transport journey opportunities.

Jacobs undertook scheme development, which included exhibitions and community consultation. In October 2005 Network Rail promoted a private bill in the Scottish Parliament, while in January 2006 the newly formed Transport Scotland became the client. In May 2006 Network Rail submitted detailed proposals for scrutiny by a Parliamentary Committee. There were only 79 objections and 39 minor changes to the plans and the Bill was approved by Parliament in March 2007.

Project scope

In the 25 years since the railway between Airdrie and Bathgate had closed, the route was still intact, but drainage, earthworks and structures had degraded. The Airdrie to Bathgate (A2B) project required a cycle route which used the track bed to be diverted. Rebuilding and electrifying the railway between Airdrie and Bathgate was the most dramatic change, but major works were required elsewhere. Seven miles of the busy “E&G” route between Edinburgh Haymarket and Newbridge Junction would be electrified, with the latter being changed from a single lead to a double junction. Track and signalling alterations and electrification were required on the Bathgate branch. This would be the largest electrification project since the completion of the ECML in 1991. In addition to five new and two rebuilt stations, an EMU servicing depot would be built at Bathgate.

Managing delivery

There were special reasons for the rail industry to perform well. In 2007 the Airdrie to Bathgate project was an opportunity to build credibility with the new Scottish National Party government and to showcase the benefits of rail investment. Early 2006, Network Rail tasked a team led by Hugh Wark, which had delivered the Scottish portion of the West Coast Route Modernisation project, with managing this new project. After joining British Rail in 1973, Hugh had passed through its excellent engineering training scheme and pursued a classic railway civil engineering career at local, Area, Divisional and Regional level. He moved to the WCRM project, under Bechtel programme management. After 33 years on Scotland’s railways, Hugh had, after, the advantage of knowing many key people in Network Rail, ScotRail and the contractors. This helped build trust and speed decision-making with this and the subsequent Borders Railway project.

Programme and implementation challenges

Work east of Bathgate could be undertaken under Permitted Development rights and formed an early part of the project. Key milestones were an intensive 16-day blockade for track and signalling alterations on the branch in July 2008 and re-modelling Newbridge Junction in October 2008.

Work on the old formation west of Bathgate was challenging, Scott Wilson had begun detailed design in October 2007 and extensive site investigation work was undertaken to inform mine remediation works and the design of new structures. The geology could hardly have been less helpful, with peat and boulder clay overlying shallow mine workings.

Not all design issues could be foreseen when the Parliamentary plans were drawn up. Where detailed design showed the need for work outwith the powers granted by the Act, the project team had to choose between a sub-optimal compliant design and the programme risk of a fresh application to the local planning authority. At Caldercruix station the proposed site was on top of an old paper mill’s toxic lagoons. In spite of difficult topography, it was necessary to revert to the old station site and to seek a planning consent from North Lanarkshire Council.

The project was a feast for bridge builders. Where appropriate, old structures were kept and repaired. However, the need to provide electrification clearances, avoid level crossings and meet modern road traffic needs led to 36 major bridge works, (excluding station footbridges). The large number of contracts managed by Network Rail and complex work sequencing over 32 miles of worksites required a sizeable client project team. Work on the new route was delayed by adverse weather. The Summer and Autumn of 2009 were exceptionally wet, with rainfall 60% above the seasonal average. Sites became waterlogged. Drier weather in December enabled some recovery, but hard frosts and heavy snow followed, again hampering progress.

Treading softly

The works under the Act were implemented in strict accordance with a project-specific Code of Construction Practice produced by the project team. This governed hours of work, noise, dust and the protection of water courses, flora and fauna. Effective monitoring arrangements were in place to ensure compliance. Considerable efforts were also made to communicate properly with local residents.

To the finishing post

Improved weather in Spring 2010 enabled intense activity on multiple fronts. Tracklaying was substantially complete by late August, followed by the OLE, and driver training commencing in October. The new service began as planned, on 10 December 2010, although two late complications arose. Delayed deliveries of new EMUs for other routes meant that some units for the new service had to be redeployed, leading to a staged ramp-up of the service. Due to wintry weather, work at three stations lasted until early March 2011.

The new services were instantly popular and revolutionised travel to work, education and leisure opportunities across the region, especially for those without access to a car. This has stimulated inward investment and house building along the route, reversing the area’s long economic decline.

Part Two – The Borders Railway

The North British Railway opened its line from Edinburgh to Hawick in 1849. The new line featured steep gradients, major earthworks and many reverse curves. After passing through mining villages outside Edinburgh, the line climbed steeply to its 880 feet summit at Falahill, before descending along the Gala Water valley to the mill town of Galashiels and then via Melrose and Newtown St Boswells to Hawick, 53 miles from Edinburgh.

A 44-mile extension of this line to Carlisle was opening in 1862. This traversed high moorland and had a 1,006 foot summit at Whitrope.

Closure

Dr Beeching’s 1963 report “The Reshaping of British Railways” proposed the whole route from Edinburgh to Carlisle for closure. An anti-closure campaign in the Borders was too little, too late. So, whilst less busy lines survived elsewhere, passenger services ceased in January 1969. This left the Scottish Borders as the only British region with no station.

Unfortunately, the formation was only protected for two years after closure. As a result, it was blocked by numerous road improvement and some housing developments and viaducts in Hawick were removed. In 1988 the Edinburgh City Bypass severed the route.

We want our railway back

The Campaign for Borders Rail (CBR) was formed in 1999. This group effectively garnered public and political support. In 1998, given the Borders’ poor public transport links, central and local Government commissioned Scott Wilson to undertake a cost / benefit study into railway re-opening. This concluded that the best option was an all-stations service, calling at new stations in Midlothian, then Galashiels and terminating at a park and ride site at Tweedbank, between Galashiels and Melrose. Following lobbying by CBR, a local authority-led group, which included Railtrack and ScotRail proposal a mixed single/double track route to Tweedbank.

Approving a Borders Railway

In 2004 the group published its business case, which included a disappointing Benefit to Cost Ratio of 1.01 to 1. The apparently weak business case gave the politicians a problem. The reopening campaign had strong support from the public and many, but not all, politicians.

In March 2005, the Scottish Transport Minister stated a conditional willingness to fund the project. The Bill Committee’s report was published in May 2006, with a parliamentary vote in favour of the project the following month.

In 2007 the project was transferred from the local authorities to Transport Scotland. The project had been strongly supported by Independent and Liberal Democrat politicians in the Borders and Labour politicians in Midlothian. Happily, the project was still backed by the incoming SNP administration, although due diligence and a spending review delayed progress. A design published in 2009 featured three long dynamic loops.

Transport Scotland wanted to explore an innovative design, build, finance, maintain, (DBFM) model. Once Holyrood had achieved Treasury clearance, an OJEU Notice was issued in December 2009 and an Invitation to Tender later issued to three consortia. Problems then multiplied. It became clear that previous construction estimates had been over-optimistic. Following the banking crash, house building forecasts fell, weakening the business case. In late 2010 one of the bidding consortia withdrew, followed by a second in June 2011.

To rescue the project, Transport Scotland abandoned the DBFM model and, in September 2011, announced that Network Rail were to deliver the project. This was to be done under the management of Hugh Wark and his project team. At the time Hugh felt that his team “has learnt a lot on A2B, all positive things we take to the Borders”. The similarities between the two projects were the management of consents, working with neighbours, commissioning a new railway, significant roadworks and utility work on the same scale as A2B.

After due diligence and the ORR’s agreement to funding via the Regulatory Asset Base, contracts were signed by the Scottish Government and Network Rail in November 2012. There were two late scope changes. The Tweedbank terminus was expanded to permit loco-hauled charter trains. However, compromises were necessary to keep within funding limits and the length of double-track was reduced from 16 to 9.5 miles, which limited capacity and posed a performance risk. However, much better a sub-optimal railway than no railway at all!

his is how you do it

After all the preceding problems, Hugh Wark, his project team and their contractors then gave a masterclass in rapid and efficient project delivery. Following competitive tendering, Network Rail quickly awarded key contracts, notably to BAM Nuttall, who had performed well on Airdrie to Bathgate and had been part of the last consortium bidding for the DBFM contract. There were many challenges:

• Major works were required to pass under the Edinburgh bypass.
• Extensive mine remediation was required in Midlothian, especially on the Shawfair deviation. Undocumented early workings and the need to maintain underground mine water flows were significant issues.
• Old structures records were mostly unavailable.
• Cuttings had been infilled at Eskbank, Gorebridge and Galashiels, and large quantities of colliery spoil had to be excavated on the Shawfair deviation.
• 42 new or reconstructed bridges and 10 km of new roads were required.
• Complex works were required to raise the level of the railway through Galashiels to avoid reinstating restrictive low under-bridges.
• Future-proofing for electrification altered the land-take at some structures and had not been allowed for in the Parliamentary plans.
• There were concerns regarding Bowshank tunnel, which had been shortened and heavily repaired in the 1950’s.
• An 80-metre span bridge was required at Hardengreen over a major junction on the A7.
• All retained structures, including 95 bridges, required repair.
• Work on the many bridges spanning the Gala Water and the bridge over the Tweed was subject to exceptionally restrictive environmental conditions.
• Extensive work was necessary to stabilise cutting slopes, especially on the long climbs either side of Falahill summit.
• The hilly terrain would complicate achieving GSM-R coverage.

However, the team also had key factors in their favour:

• There was a strong collective will to succeed.
• Key team members knew and trusted each other, having worked together on previous projects.
• Network Rail’s Asset Engineers were supportive and pragmatic, (setting a target of no major works within 5 years of reopening).
• BAM Nuttall had a sound grasp of the project.
• Early contractor involvement enabled an efficient target cost to be developed, based on a sensible allocation of risks.
• Transport Scotland agreed to retain some unquantifiable risks.
• Many old bridges were in surprisingly good condition.
• Logistics risks were mitigated by creating a rail supply depot at Millerhill.
• The three local authorities were very supportive.

Early contractor involvement soon brought benefits. BAM Nuttall demonstrated the advantages of reinforced earth techniques for retaining walls and wingwalls, of concrete rather than steel bridges and of on-site quarrying at Falahill. They were also very adept at reusing material within the works, minimising off-site disposal.

Delays arose from unexpected sources. The diversion of a single British Telecom fibre cable caused a lengthy delay at one site. At another, designs had to change due to a local authority oversight, which prevented a planned road closure. “Blowing” the signalling fibre cables through ducting, (a UK rail first), was very tricky, because siteworks had prevented all the ducting being laid at the same time. Long lead times for internet connections for ScotRail station systems were also a concern.

Additional unhelpful industry bureaucracy meant that the challenges were not just practical. New interoperability regulations prevented the use of 30,000 good quality used F27 concrete sleepers, which had been carefully stockpiled. It got worse! Many meetings were required before the ORR and the Network Certification Body would accept that the standard G44 sleeper complied with the regulations! The new CSM risk assessment process had to be back-fitted to a project whose design and early works had begun many years previously!

An otherwise excellent safety record was unfortunately marred in November 2014, when an item of plant malfunctioned, causing a concrete sleeper to fall and crush a worker’s leg. Track work was immediately halted, while the root cause was determined and corrective measures implemented.

Key milestones were:

• February 2014 – craning-in the main beams of Hardengreen viaduct.
• May 2014 – completing the works on the Edinburgh bypass.
• August 2014 – completing repairs to the many bridges over the Gala Water,
• September 2014 – completing structural repairs and installing slab track in Bowshank tunnel.
• February 2015 – completing track laying.
• June 2015- signalling and GSM-R commissioning.
• August 2015 – completing driver training.

At Galashiels, a new supermarket left only a narrow corridor for the railway. In spite of this constraint, Scottish Borders Council created an excellent bus-rail interchange to maximise public transport benefits.

Working with local communities takes many forms. In the Scottish Borders rugby is a serious matter. In Galashiels, new bridges were painted maroon to match the local club’s colours. The standard Network Rail dark green looked dangerously like the colours of arch-rivals Hawick!

Opening

Passenger services began on 6th September 2015. After 46 years, the Borders were re-connected to the network. Three days later the line was officially opened by Her Majesty the Queen, with Scotland’s First Minister in attendance. This memorable event made national and international news.

Postscript

The Borders Railway is the longest re-opened line in Britain. Huge credit should go to the campaigners who lobbied for it, the politicians and civil servants who backed it and the railway engineers, project managers, contractors and operators who finally delivered it. It is a remarkable collective achievement, which has transformed public transport in Midlothian and the Borders.

However, services were immediately overcrowded, greatly exceeding demand forecasts for Stow, Galashiels and Tweedbank. The large car park at Tweedbank was so busy that Scottish Borders Council soon added 100 extra spaces. The shortening of the dynamic loops made punctual operation difficult. Network Rail are developing options for increased capacity at the north end of the route.

Part Three: issues and lessons

Making the case – forecasting demand

Forecasting the use of a new service is fundamental to the business case. Standard models reliably predict the effect of changes to existing transport networks. They are not good at predicting the use of a completely new service. This was a problem with the Borders Railway, where demand from Tweedbank, Galashiels and Stow was grossly underestimated, making the business case seem weaker than it really was. Forecasting techniques should be sensitive to the local circumstances and funders should welcome more accurate, bespoke methods.

Defining scope

The planning process rightly protects and compensates landowners and affected residents. In addition to property purchase, this can require more extensive works than are simply required to build a railway. When combined with the ban on new level crossings and highways works being constructed to modern standards and needs, the cost of “non-railway” works can be high.

When planning new services, there are complex trade-offs between investment cost, route capacity, resilience and the impact on other present and future rail services. This is especially tricky when trying to avoid “loading” a high cost onto a single project. Network Rail and its consultants must understand the strengths and weaknesses of their operational modelling tools and properly interpret their outputs. Drivers, signallers and other front-line staff have invaluable knowledge and can contribute greatly to scheme development. Their knowledge is not always properly used.

Timescales and completion dates

It took 7 years from the “in principle” political decision, to trains operating between Airdrie and Bathgate. This was quite fast, given the scale and complexity of what had to be done. The evolution of the Borders project was slow and tortuous, with 15 years between the Petitions Committee meeting in Galashiels and trains running. At times, the coalition of political interests supporting the project was under severe pressure.

In both projects, completing detailed design and implementing physical works were quick and efficient. The only comparable UK project is the Oxford to Bletchley portion of East West Rail, which has also had a long history, (and an easier task, with a flat route that had not been breached). The 34 months from the agreement between Network Rail and Transport Scotland to public services on the Borders Railway was superb.

There is often early pressure to commit to a completion date. With the Airdrie to Bathgate project, the team calculated that commencing services in December 2010 was achievable, but this became very challenging. The issues were well managed, but the commitment to the end date had been premature. With the Borders Railway, Network Rail were right to resist pressure from Transport Scotland for a much earlier completion date.

Management

The roles funding stakeholders, Network Rail and contractors are different and should not be confused. Efficient project delivery relies on clarity on what is to be delivered and the respective roles of the parties involved. With both these projects, these lines were not blurred and this helped all parties to achieve their objectives.

Hugh Wark and his team were ideally suited to the task. Hugh was respected by Network Rail colleagues and contractors alike. His team could deal with complex engineering issues and intricate project dependencies. Some believe these roles can be outsourced from Network Rail to contractors or specialist project management providers. I believe this is misguided. Being able to change and develop the network cost-effectively must be a core skill for Network Rail and GBR. Despite huge pressure on overheads, they must develop and retain individuals with the skills and experience to enable the organisation to be an effective, informed client.

Teamwork is often taken to mean collaboration between Network Rail and its contractors. This is crucial, but just as important is the relationship between project teams and others in Network Rail, (asset engineers, operational planners, asset maintainers, etc.). Both projects were a superb team effort by Network Rail as a whole. Sadly, this is not always the case. Its absence leads to delay, higher costs and sub-optimal outcomes. Close collaboration is clearly also needed with funders, TOCs and FOCs.

Contracting

Procurement legislation is complex. Invitations to Tender must be expertly crafted and assessing bids properly requires great care. This needs close collaboration between the technical specialists and procurement professionals. Poorly drafted ITTs or inexpertly assessed bids can cause huge problems.

In 2006, Network Rail’s Infrastructure Projects policy was for fixed price contracts, hence their use on Airdrie to Bathgate. This led to unhelpful friction between client and contractor. Strong personal relationships and a collaborative ethos kept matters on track, but a better approach was possible. With Borders Railway, a mature form of target cost contract was used, with risks allocated to those best placed to manage them. This incentivised the contractor and achieved a high degree of financial and programme efficiency, with no loss of client control. Alliance contracts can work too, but they do not avoid the need for Network Rail to be a strong, knowledgeable and decisive client.

Achieving value for money

Both projects cost more than £300 million (around £450 million at today’s prices). There are many ways to measure value for money and efficiency, but in both cases the Network Rail project team kept a vice-like grip on scope, cost, and programme. They chose their consultants and contractors wisely, got the best out of them, used their expensive resources efficiently and maintained a robust delivery plan. Re-design was minimised. None of this happened by accident. Hugh Wark provided strong, expert leadership and, crucially, an appropriately-sized Network Rail project team of competent, experienced individuals supported him.

What makes the cost of the Borders Railway so impressive is the scale of the “exceptional” items. With extensive mine remediation works, complex works at the Edinburgh bypass and major earthworks, major bridge and highways programmes to deliver, I believe forensic analysis would demonstrate that this and the Airdrie to Bathgate project are among the most cost-effective major projects delivered since railway privatisation.

Network Rail has rightly been criticised for the cost of investment projects. These two projects predate its Project Speed, PACE initiatives, and show the problem was not universal and that, in some cases, high levels of cost-efficiency were achieved. Given the scale and complexity of the works, the circa £15M per mile for Borders Railway sets a very high standard.

My sincere thanks to those who assisted in the preparation of this article. The views expressed, and any errors, are mine alone.

Peter Heubeck joined British Rail in 1979. After front-line roles he became Railtrack’s Zone Commercial Manager in Manchester and he played a leading role centralising and modernising engineering and track materials supply. He later established Trackwork Moll, a major sleeper manufacturer, and now works as a consultant. [email protected]

This article appeared in the January 2023 issue of the PWI journal and is reproduced by kind permission of the PWI

Image credit: PWI