This is my external view of the TfW led procurement for the Metro and Wales and Borders franchise in 2016-2018; and my insight from working with MTR on their bid. Then a brief reflection on the efforts of TfW, their staff and supply chain from 2019/2020 through 2024 to actually build and operate the Metro.
Those closer to the procurement engine and decision making and the later implementation teams may tell a better and/or different story?
In this chapter I cover:
- 6.1 Metro procurement 2016-2018
- 6.2 The MTR/BAM Nuttal joint venture team and bid
- 6.3 The Keolis Amey solution for Core Valley Lines
- 6.4 Building the Metro 2019-2026
- 6.5 Gaps & so still work to do…
- References
6.1 Metro procurement 2016-2018
After what seemed liked years either haranguing or working with/for Welsh Government, by Spring 2016 I was out.
As I left to take up a part time role at Cardiff University, my ex-colleagues at Transport for Wales (TfW) were busy preparing for a procurement process that would be much more complex and time consuming than previous DfT rail franchising procurements. The addition of the Metro and the anticipated transfer of the Core Valley Lines (CVL) from Network Rail (NR) to a new infrastructure body, and a competitive dialogue process would take nearly two years to conclude. The mode agnostic approach as regards the Metro also included another procurement innovation (at least as far as rail franchising was concerned). Typically, in rail franchising, the government sets out what it wants and asks bidders for the best price/least risk solution to deliver its service requirements. In this instance, what Welsh Government via TfW did was set out its high-level requirements and the capital and revenue envelope it could afford, and then asked the bidders to develop a solution that satisfied those financial limits within an acceptable risk profile.
The procurement also I think unhelpfully combined the Metro with rail services for the entire Wales and Borders franchise. This though was probably unavoidable given the provisions of the rail act and the fact the CVL was still part of the national rail network.
So, I think we have to applaud Welsh Government for taking on and running, via Transport for Wales, the biggest procurement ever undertaken by WG and the most complex in the rail industry since privatisation. The amount of analysis and diligence displayed in that process was immense, requiring different teams to engage with complex details across multiple workstreams; these covered customer service, stations, ticketing, fares, infrastructure management, operations, rolling stock, timetabling, HR, implementation, etc.
This is all the more remarkable, considering that TfW did not exist until 2015. We should also not underestimate the importance of the political support and empowerment of officials provided during this process.
One undesirable feature of the more complex process, was the time and effort required to run it and to bid. During that process four bidders initially engaged and expended a significant amount of resource and energy. I estimate each bidder will have invested between £5-10M in preparing and submitting their bids as part of a process that took nearly two years. I also think that no matter how objectively structured and quantitative one’s procurement process is designed to be, there is always room for subjectivity, different interpretations and so scoring.
For me, the high-level vision and objectives of the South Wales Metro had been developing since my first report was published in 2011; these were, perhaps, best summarised in “Rolling out our Metro” published to launch the project in November 2015. A more formal definition of WG’s priorities for the Wales and Border franchise[1] (inc. South Wales Metro) was published in March 2017 as part of the procurement process and to reflect the consultations undertaken. For me, some of key features of the Metro vision were, embedded in the procurement:
- Faster and more frequent services – the need for all places to have at least four services an hour and to reduce journey times between Cardiff and the Heads of the Valleys (HoV)
- Direct connections to both Cardiff Bay and Central from the Heads of the Valleys – it was this requirement that would later, given the chosen solution, become a bit of a millstone
- Extendibility – to enable future extensions and/or add stations to serve those large/dense population centres currently disconnected from the rail network in SE Wales
In my naivety, given the underlying requirements which I had shaped, I fully expected a Light Rail (LR) solution of some form to win the bid. Not least because I “knew” that to affordably deliver the comprehensive 4tph requirements set out would require a de-designated LR network using a Tram Control System, which would enable more inherent network capacity and operationally flexibility than a traditional HR block signalling system. It would also enable a network and services, given DOO, with the potential for lower unit costs and more operational flexibility. This was to precipitate a lesson for me about the need to challenge one’s assumptions on a regular basis.
In December 2016 four groups were working on bids to build and operate the Metro: MTR/BAM Nuttal, Keolis Amey, Abellio and Arriva. However, by early 2018 this was down to two, The MTR Corporation/BAM Nuttal joint venture and Keolis Amey
6.2 The MTR/BAM Nuttal joint venture team and bid
Aside from my new role at Cardiff University, I was approached by the MTR Corporation[2] and in July 2016 joined them (again on a part time consultancy basis) to support their bid to operate the Wales and Borders franchise and to develop, implement and operate the South Wales Metro. The formal bidding process run by TfW started in the summer of 2016; I wanted to help MTR put in the best bid, which I felt would set a standard that the others would have to beat. I would learn a lesson (again) on this journey. A recurring theme.
The team established by MTR, in a joint venture with infrastructure partner BAM Nuttal[3], included HR and LR expertise with transport planners, operators, rolling stock expertise, etc from the likes of Arup, SNC Lavalin, and others. The team prepared an outline solution which was submitted in the December of 2016[4] then MTR were invited to develop and submit a full bid through to December 2017.
It was during the 2017 that I accompanied my MTR colleagues Elizabeth Yeung and Carl Waring on a trip on the Valley Lines to Treherbert. Elizabeth had never been to Wales, had no idea of the industrial history or topology of Cardiff and the valleys, so had no pre-conceived ideas about the place or its potential.
Elizabeth found the Rhondda valley and the urban/green contrast beautiful; amazed at wind farms next to rows of mining cottages at Treherbert, found the legacy of glaciation etched in the hillside stunning. We visited Pontypridd which she felt was a wonderful place with a magnificent station, authentic high street and market, and a fantastic river (with fly fishing – the salmon were jumping). She felt a real sense of a place with an incredible industrial, cultural, environmental and geographic legacy.
The lesson for me is that people from the outside often see our region in a more positive way than we do ourselves. They are unconstrained by our nostalgia for an industrial past that will never return and see the unique urban & green environment in very positive way.
I am optimistic the Metro can be the enabler of agglomerative benefits and a catalyst for wider community regeneration and the development of local economies all across southeast Wales.
By the end of 2017, after a year of further detailed work, MTR’s final bid for the Metro was completed. It set out a fully segregated and de-designated, “non-mainline”, Light Rail network “cut off” from the NR network south of Cardiff Central. The proposals set out a single fleet of high floor trams (like Manchester Metrolink, although with a crash worthiness profile that would allow them to mix with HR and freight and so enable subsequent update for tram-train operations) offering a minimum of 4tph across the whole network, with reduced journey times and double the capacity (with room for more). MTR also developed a solution to allow the infrequent freight services to continue from Cwm Bargoed (there is precedent for freight operating on “non-mainline” railway systems). The proposal also provided for easier and lower cost expansion.
At that stage MTR proposed an electrified 750V DC solution. The only “downside” was perhaps that through services from the Vale of Glamorgan (VoG) Line to the valleys would, for a period, be severed and there was no initial remedy for the 2tph on the city Line which would remain HR (although MTR did set out an extension option to convert it to LR). The crash worthiness specification of the tram rolling stock was to retain the future option of fitting out for through running as tram-trains onto the VoG and NR kit. Whilst there were project risks, these were known; no new engineering innovations were proposed. In fact, everything proposed by MTR had been shown to work elsewhere. The system would look something like the Manchester Metrolink network.
MTR had chosen (rightly in my view) to minimise the engineering and deliverability risks in their bid, but in so doing had perhaps introduced more institutional and Industrial Relation (IR) challenges and the requirement for more of a “big bang” Light Rail conversion.
However, it was during 2017 when there were still three bidders, and the preparation of the final bid, that some alarm bells started ringing for me. I was on the outside, so I’ll never really know what happened or what decisions were made, but for some reason, the requirement for 4tph on the Coryton line was dropped by TfW to 2tph as were direct services from the Rhymney Line to Cardiff Bay. This for me suggested that WG/TfW were providing room for a more traditional HR block signalling system – and removing one of the “safety valves” I had originally included in the scope back in 2015. It seemed to me that something had spooked WG. I also suspect that some of the rail unions were not well disposed toward a de-designated “LR” network as it could open the door to changes in staff T&Cs and working practices (e.g., Driver Only Operation – which is normal for LR systems and LRVs). Having said that the procurement did set out there would need to be a second member of staff on all trains.
The MTR bid for the Metro scored as I recall very highly – in fact I believe higher than the winning Keolis Amey bid for the Metro given it provided more CVL capacity and services. However, the KA bid scored higher when viewed across the whole W&B franchise. I suspect (and maybe I’ll never know), that what WG and TfW did, was lean to avoid the institutional and industrial relations risks of a full LR solution and instead take on a solution with what I considered, more engineering risk, and in the longer term, higher costs (both capital and opex). I will probably never know what really happened.
I recall James Price calling me in April 2018, to tell me what we all already knew on the MTR bid team. The winning bidder Keolis Amey and their solution were announced in (although Keolis would eventually step away because of Covid revenue issues in 2021[5]). This is, for the most part, the solution and timetable TfW are now building and planning to operate into 2025. Whilst I applaud the ambition and scale of the process, the winning solution, for me, contained risks and challenges that I felt then could add to capital costs and deliverability. The solution would also “bleed” HR thinking and costs into TfW operations and inhibit plans for further service expansion. But it was an innovative solution and one that probably navigated the institutional and Industrial Relations (IR) risks better than the MTR one – as well as unpicking some of my earlier assumptions from my work in 2014/5 with WG.
6.3 The Keolis Amey solution for Core Valley Lines
Despite my reservations, the KA solution played the institutional constraints and changing requirements very well and set out a very innovative solution – so credit to their bid team led by Colin Lea (now playing a lead role for TfW and supported by the likes of Andy Holder, Steve Whitely and others). The KA solution was to be based on retaining “mainline” designation (apart from the bay line), a new Siemens signalling system that would actually enable a little more capacity through Queen St north (again unpicking one of my earlier assumptions), new discontinuous 25Kv electrification and new electric bimode and trimode rolling stock. As we had set out in 2015, the solution would also require the transfer of the CVL from NR to WG. Although I was and still am not clear what the point of that is if you are not going to “de-designate” the network?
The Core Valley Lines (CVL) Transformation is now a £1.1Bn[6] (up from £734M anticipated in 2018) investment that will deliver “turn up and go” services of at least 4tph to Merthyr, Treherbert, Aberdare (using new Stadler City-link Tram-train “Metro Vehicles” Figure 73 ) and Rhymney (using new Stadler HR “FLIRT” Trimodes Figure 74). There will also be 5tph (Trimodes) to Barry with 2tph operating on the Vale of Glamorgan (VoG) line to Bridgend. The network will also include new Stations at Crwys Road and Butetown north.
This project will double the capacity of the valley line network, reduce journey times and provide a real alternative to car use across much of the Cardiff Capital Region. In addition, the recent announcement (Jan 2023) of funding will enable the first phase of the Cardiff Crossrail between Cardiff Central and the Bay Line and hopefully onto Pierhead St, to be delivered.
However, the combination of the discontinuous OLE solution and new battery powered stock had never been tried before together in this way. I also think that we really only needed one new type of RS for network – a tram-train; not two, both of which need batteries adding to their weight) I did set out in a more circumspect manner my assessment of the process and some of the risks of the proposed solution in my evidence to the Senedd Committee in December 2018[7]
Figure 73 Transport for Wales Stadler Citylink Tram-Train Metro Vehicle
Figure 74 Transport for Wales Stadler FLIRT Tri-mode train
Figure 75 Winning KA Bid for Metro (from presentation to MSs in 2018)
My assertion, which I stand by today (and others may disagree or tell a different story), is that procurement was compromised to a degree by official, trade union and political influence and a lack of LR knowledge in the procurement team. I still maintain the MTR Metro solution was technically the best and would have enabled ongoing lower cost operations – but was for other reasons probably undeliverable.
However, we are where we are, and I know that the route to the top of the mountain is never straight, especially when you look down from the summit. We do have a very good solution that will provide the foundation for network growth and expansion into the 2030s.
6.4 Building the Metro 2019-2026
I cannot understate the scale and complexity of the challenge of building the South Wales Metro. Since 2018, TfW have been undertaking a major upgrade and enhancement of what is an operational railway, a Victorian railway that, let’s be honest, has had very little investment since it was opened to move coal down to Cardiff from the valleys in the 19th Century. The number of previously unsighted issues and liabilities should not be underestimated.
Figure 76 Taffs Well Tram-train depot Spring 2024
The integrated effort of teams working on design, trackwork, permanent way and civil engineering, electrification, signalling, rolling stock, stations, timetables, operations, rolling stock, branding, marketing, stakeholder engagement, project management, asset management, etc is frankly staggering and to most people, completely invisible. I have seen some of their work, and the pressure they all face given the expectation, but James Price and his team[8] are an energised, committed collective of multi-disciplinary professionals all focussed on delivery. One also has to acknowledge the contribution of the supply chains via TfW Rail Services, Amey Infrastructure Wales (AIW) and their major contractors like Amey, Griffiths, Siemens, Stadler and CAF amongst others.
The potential longer-term programmes set out later, will build on the current contracted project to enhance current Wales & Borders rail services. This will result through 2024/2025/2026 in additional rail services, £800M of new rolling stock for the entire Wales and Border services (which includes the new stock for the South Wales Metro) , and the completion of the current phase of the South Wales Metro via the transformation of the core Valley Lines[9]. That work is well advanced as the new tram-train depot emerging at Taffs Well demonstrates Figure 76.
The reality though, for most people, is “seeing is believing”. I was quoted in January 2021 as saying that even my mother didn’t believe the Metro would happen[10]; this appeared as a front-page headline in the Western Mail Figure 77.
Figure 77 Western Mail Headline January 2021
Actually getting on a new tram-train sometime late in 2025 is what will make the difference. In fact, I sense that when people actually see it operating there will be a palpable sense of disbelief. Perhaps not at the same scale, but I hope they will also be as impressed as people were (including me) in London when the Elizabeth Line opened in May 2022 (over 39 years after it was first announced.).
To exemplify, when TfW launched one of the new Stadler FLIRT trains on the Rhymney Line at Caerphilly earlier in 2023, waiting passengers had to be told they could actually get on. We have become so used to old unreliable rolling stock, that when something new and shiny turns up at a station, many people assume it must be for someone else.
I suspect passenger numbers on the Core Valley Lines will exceed expectations when it is part of a fully operational “Metro” in 2026. As a comparator, the latest patronage figures from the DfT[11] for the Manchester and Tyne and Wear Metros are instructive. The Manchester Metrolink is approximately 100Km in length with 99 stations and has annual PAX of 36M; Tyne and Wear Metro is a little shorter at 80Km with 60 stations and has a PAX of 30M. The CVL part of the South Wales/Cardiff Capital Region Metro is 170Km with 56 stations and pre-Covid handled about 10M PAX. In fact, the catchment populations of the CVL and T&W Metro are about the same, so I see no reason not to expect the CVL PAX figure to at least double by 2028/9; However, to ensure we do, I anticipate we will need more capacity sooner rather than later and will certainly need address some of the key omissions in the current programme.
Figure 78 Upgrade WIP at Cardiff Bay station August 2024
So, from Cardiff Metro heresy in 2011 to South Wales Metro reality in just over 13 years. That is actually pretty impressive given how long most UK transport projects take to gestate and deliver. For example, London’s Crossrail was actually first announced when I was living in London in 1987; the current incarnation initiated in 2009, opened in May 2022, four years later than originally planned and its costs have gone from £14Bn to about £22Bn. The Tel Aviv Metro’s new 24 Km red line[12]was first conceived in the 1990s, and after much development, construction started in 2015. It started operating in 2023, over two decades after first being announced in 2003, rather than the six years intended; and will likely cost well over $4bn to complete. There were also major cost and time overruns on the Great Western Electrification Programme[13] and now HS2. In Wales we have also seen delays and costs balloon to over £1Bn for the completion of improvements to the Heads of the Valleys Road between Hirwaun to Abergavenny.
Worth also reflecting that since London’s Crossrail opening, no one is talking about the cost overruns and delays…it’s all about how amazing it is.
So, lets applaud what we are now delivering and turn our attention to what happens next.
6.5 Gaps & so still work to do…
In setting out and developing the next phases of the Metro to 2030 and beyond (and much work already in progress) we have to acknowledge some of the key gaps and constraints[14] (as at September 2024) of the current and committed changes to network and services. These will need to be addressed in future phases (See 15.5 Cardiff, the Cardiff Capital Region and Crossrail). These include:
- Only 2tph on both the City and Coryton lines – which presents a very imbalanced offer for Metro services in Cardiff and an under provision in many parts of the city that given catchment populations also need a minimum of 4tph.
In fact, unlike everywhere else on the CVL, over half the stations in Cardiff will see no improvement in service frequency and so will not really be getting a “Metro”. I briefly cover the primacy of service frequency over journey times for short urban trip in 12.5 Service frequency
The focus on ensuring sufficient capacity at Llandaf station for Metro services – whilst ignoring most of the other stations in Cardiff is a little odd. If you ran ten shops across a city, you’d want to ensure they were all performing to their maximum – not just one of them? - Further CVL stations are required at places with large local catchment populations or near trip generators. These include Roath Park, Gabalfa, Ely Mill, Treforest Industrial estate station relocation (to Nantgarw) and Pontypridd north, as well as a major upgrade at Cogan with the introduction of a platform on the Penarth branch to enable interchange between VoG line and Penarth services
- Simplifying an unnecessarily complex and less legible service pattern, linked to the European Funding Targets (ERDF) targets originally set out in 2015 (partly by me.) (so, a requirement to have services from Merthyr, Treherbert and Aberdare route to both Cardiff Bay AND Cardiff Central – the requirement to serve the bay from the Rhymney Line had already been dropped during procurement) (See 12.7 Network legibility & simplicity)
- The exclusive operation of Tri-modes on the Penarth and especially the Coryton branches constrains potential extendibility (esp. the Radyr/Coryton “Circle” link); both these lines are much better suited to LRV operation
- The retention of guards on all tram-train services will add to operating unit cost and makes some extensions and service expansion opportunities less affordable
- Two types of CVL rolling stock adds to operational complexity and inhibits flexibility
Beyond the scope of procurement there are also wider issues to address:
- There is no electrification on the VoG Line, Penarth branch or SWML west of Cardiff which would enable great operational flexibility and more diesel free operation (eg Tram-trains to Penarth)
- Similarly, whilst I appreciate the engineering requirements and constraints of the 25Kv solution, the bay line really ought to have been catenary free (so no overhead electrification) or used a 750vDC solution. There will continue to be urban realm constraints without this measure, for example to remove more of the Bute Street wall
- There is no real rail/bus integration re: network design, timetable, fares or ticketing; something WG’s planned bus reform will help address
- The wider NR network has infrastructure constraints which limits service frequencies on the Marches, Ebbw Valley and Maesteg Lines
- I really think that de-designation to non-mainline/tramway needs to extend beyond the Cardiff Bay line.
So yes, today, we are grappling with some legacy challenges as set out above. Nonetheless, the route we have come was in reality probably the only route open to us.
Nor would I change it, we are where we are and move forward from this point. I would also like to re-iterate, despite my critique, what an amazing job has been done to move this project through procurement design and now into implementation – just too many people to name them all.
It should also be said that Metro would not have happened without the commitment of Welsh Government and the actual existence of a devolved administration in Wales; and to be clear there was no prospect of UK Government leading and/or providing the majority funding for this project, despite this still being a UK Government responsibility.
Fundamental to the long-term success of Metro, was the inclusion of the concept in the 2011 Labour Senedd manifesto thanks to Mark Drakeford, followed by the engagement of my work by Edwina Hart, and the ongoing support of FMs Carwyn Jones, Mark Drakeford, Vaughan Gething and now Eluned Morgan (who was Chair of the Cardiff Business Partnership in 2013) and key Cabinet Ministers including Julie James, Lee Waters and Ken Skates. Despite sometimes a challenging relationship with me, the role of the region and its local authorities (and people like Andrew Morgan, Peter Fox, Russell Goodway, Brendan Toomey, Bob Bright, Huw David, Huw Thomas etc), and their support of the Metro was also fundamental.
I suspect nobody got everything they wanted, but everyone got some of what they wanted.
I think Lee Waters deserves special broader mention given he has led from the front re: transport decarbonisation and implementing some very challenging policy such as 20mph; he was also the leading forced behind the earlier WG Active Travel Act. History, I suspect will look very favourably on his contributions.
References
[1] Welsh Government (2017), Priorities for the Wales and Borders Franchise Rail franchise and Metro: policy priorities
[2] The MTR Corporation is a Hong Kong based company established in 1975 as the Mass Transit Railway Corporation with a mission to construct and operate an urban Metro system to help meet Hong Kong’s public transport requirements. It has expanded operation internationally and is the current operator of London’s Crossrail. MTR > Our Business
[3] BAM is a UK based construction and property services company | BAM
[4] Hats off to people like Jeremy Long, Neil Walker, Clare Parmenter, Roddy Beynon, Peter Cushing, Mike Bagshaw, Gwyn Roberts, Katie Allister, Russell Cohen, Chris Wood, Chris Clark, Rob Needham, Oliver Bratton, Mark Beckett, John Fagan, Dan Jones, Huw Jones, Carl Waring and many other – it was a big and very capable team
[5] Welsh Gov 2021, Written Statement: New Delivery Model – the future of rail in Wales and the Borders
[6] TfW, March 2023, South Wales Metro final cost re-evaluation (tfw.wales)
[7] Mark Barry, 2018, Evidence to Senedd Committee Article (senedd.wales)
[8] An energised and capable team: Led by TfW CEO James Price and including Karl Gilmore, Dan Tipper, Heather Clash, Alexia Course, Anthony McKenna, Marie Daly, Jan Chaudhry van der Velde, Colin Lea, Steve Whitely, Andrew Gainsbury, Phil Rawlings, Nathan Sealy, Gareth Pagett, Aaron Silva – and many others….including at AIW TfW Rail and key supply chain partners
[9] Transport for Wales, Core Valley Lines Transformation | Transport for Wales (tfw.wales)
[10] Mark Barry, 2021 Blog Do you believe in Metro? – Mark Barry (swalesMetroprof.blog)
[11] DfT 2023, Light rail and tram statistics (LRT) – GOV.UK (www.gov.uk)
[12]. Bloomberg, Aug 2023 Tel Aviv Light Rail Opens Red Line With Stops From Bat Yam to Petah Tikva – Bloomberg
[13] House of Commons, Public Accounts Committee 2017, House of Commons – Modernising the Great Western Railway – Committee of Public Accounts (parliament.uk)
[14] See various M Barry blogs and Senedd paper:
Metro, subsidies & Cardiff… Some tough choices and compromises? – Mark Barry (swalesMetroprof.blog)
Metro Moans for 2023… – Mark Barry (swalesMetroprof.blog)
Cardiff West Junction… – Mark Barry (swalesMetroprof.blog)
Submission to the Economy, Infrastructure and Skills Committee, 2018 Professor Mark Barry.pdf (senedd.wales)