High construction costs for rail in the UK - causes and solutions

[eldomtom2]

I believe the high cost of rail construction projects is a key issue of railways in the UK, and I feel it is underdiscussed. I shall start by giving two examples of the problems caused by high costs.

Network Rail is already talking about the possibility that lines may have to be abandoned due to extreme weather - and this is fundamentally a cost problem. If construction costs were cheaper, it would be easier to justify reconstruction.

The cancellation of HS2 Phase 2 was also fundamentally a cost problem. The less it cost, the better the argument for constructing it and the worse the consequences for Sunak if he cancelled it.

But are costs in the UK higher than in other countries, and if so what can be done about? There are two main sources that I shall discuss here that answer these questions. Neither are focused on the UK specifically, but they provide good evidence for the claim that rail construction costs in the UK are high and they also discuss possible solutions.

The Transit Costs Project was performed by the New York University's Marron Institute. It conducted an in-depth survey into the causes of differing project costs, focusing especially on the US, Sweden, Italy, and Turkey, but with research done into many other countries as well.

A side note on the Transit Costs Project - Alon Levy, one of the writers, is crankish at the best of times and has cultivated a community of extremists on their blog. But their colleagues on the report seem at least somewhat more reasonable and I've never heard anything said against the report's actual figures.

The Transit Costs Project produced two databases of transit projects and their costs. The first focused on urban transit projects, covering 883 projects from 59 countries from the 1980s to today. Crossrail was the twelfth most expensive project on this list, at a whopping 1142 million 2023 USD per kilometre, only beaten out by projects in the US, Singapore, and Hong Kong.

The second database produced by the project was on high-speed rail. Here the UK has the dishonour of having the project - HS2 - that is far and away the most expensive, at 231.9 million USD per kilometre (though the dataset does not include things like projects in the US, which may be more expensive). The project provides some helpful visualisations of the data which show that no matter how you sort it - by amount of tunnels, by amount of bridges, by GDP, by time taken to construct - HS2 remains far more expensive than any of its peers.

The other source is a webinar presented by Paul Lewis, a consultant with Deutsche Bahn's US consultancy, on "Global Best Practices for Cutting the Cost of Building High Speed Rail".

His figures show the UK as being even worse than the US on the "primarily tunneled urban transit project" front - though probably this is because he averages out the cost per mile among different projects. He also finds high high-speed rail costs in the UK, though he does find them cheaper per kilometre than Dutch projets.

If there is a problem, what are its causes and how can it be solved? Again I shall rely on my two main sources here, though obviously I can't speak to their specific applicability to the UK - that's what I started this thread to discuss, after all.

The Transit Costs Project finds a large number of causes for high project costs:

• Political micromanagement and reliance on elected or politically appointed officials to make decisions rather than professionals
• Lack of internal capacity in transit agencies and the civil service, leading to a reliance on consultants
• The use of construction manager/general contractor rather than design-bid-build or design-build and the use of fixed-price contracts rather than itemised ones - the report is extremely scathing about this
• Lack of transparency when itemised contracts are used
• A focus on minimising cost overruns rather than the total cost
• Lack of coordination between different agencies
• The use of bespoke designs rather than standardised ones
• An unwillingness to use cheaper but more temporarily disruptive construction methods like cut-and-cover

(There is also a fair bit about labour costs, but the report specifically marks them as US-specific and not applying to other countries with high construction costs like the UK)

The report does not mention the UK much - though when it does it is pretty much always negative as an example of how not to do it - but it does claim that London Underground construction costs rose after changes to procurement in the 1990s reduced the role of the civil service and increased the role of consultants.

The Transit Costs Project makes a large number of recommendations, primarily focused on the US but presumably applicable generally:

• Find champions who will "advocate for the project, help with intergovernmental agreements, hold agencies accountable for budgets and schedules, and support agencies if there is political controversy"
• Import foreign experts to get a fresh perspective, adopt foreign standards rather than limiting options to what has been tried locally, and develop long-term relationships with foreign peer agencies
• Collaborate and work with local agencies, especially utilities, and ensure utilities provide clear and up-to-date information about infrastructure that may be affected by construction
• Create a culture of transparency where information about blueprints, itemised costs, public contracts etc. is freely shared with the public
• Use design-bid-build or design-build rather than construction manager/general contractor
• Ensure there is adequate in-house expertise to supervise the contractors
• Use publicly-viewable itemised contracts rather than fixed-price contracts
• Ensure the change order process is flexible and lawsuit-free
• Keep the risk in the public sector
• Award contracts based on technical merit and not just price
• Standarise regulations nationally
• Limit contingencies
• Standardize designs so that they can be repeated between different expansion projects
• Use cut-and-cover rather than mined tunnels and stations whenever possible
• Build simple station boxes, with little more space than is required for passenger circulation

Paul Lewis' webinar consists of ten main points to reduce costs:

1. Build simple and useful projects
2. Break up civil works contracts and set itemised costs
3. Manage more risk on the public sector side
4. Empower staff to engage the community (and say no)
5. Proceed with the more disruptive timeline
6. Finish planning process before environmental review
7. Boost internal staffing
8. Set agreements with partners early
9. Address institutional governance
10. Learn from peer country examples

Clearly similarities can be seen with the Transit Costs Project's recommendations.

So, what do you think of these recommendations? Do you think they are applicable to the UK and would reduce costs? What other methods not listed could be used to keep costs down?

 

[Trainbike46]

A few things that are relevant to consider are:
- What is and isn't included in a project? for example, HS2 includes significant station works, as well as lines within cities, whereas HSL-zuid in the Netherlands didn't include any stations or lines in a city
- repeatedly changing what is to be build is never good for costs (e.g. the infitely changing set of TransPennine projects)
- Our toxic politics in the UK definitely contributes to the problem

 

[mike57]

@eldomtom2 An interesting post.

To pick up on a few points

Ensure there is adequate in-house expertise to supervise the contractors

Other areas of the public sector have been taken to task in the past over this. The regulator in a safety critical undertaking actually insisted improvements were made in project delivery. The problem is often finding people with the required knowledge and experience, they have to be knowledgable enough to challenge things, I have performed that role in the past and its not easy. In our area of operation its refered to as an 'Intellegent Customer' role. I would work closely with the commercial manager to ensure that we were getting the best value for money, and that technical solutions were in line with our requirements.

Award contracts based on technical merit and not just price

This seems to be an endemic problem in industry generally, and recent procurment in the rail industry has fallen foul of this, thinking Mk5s for TPE for example, the plethora of incompatible microfleets, and poor quality carriage interiors.

The other major UK specific problem has to be political meddling, with rail projects ending up as a political football, the DfT mircomanagement and the failure of the franchise system.

 

[SynthD]

Of the 25 points, only 2 are rail specific. I hope those 23 points are considered across the board and become rules for politicians.

The first of the two rail specific points - cut and cover tunnels - is often still a political choice. Kenilworth has a large cut and cover tunnel in an urban area for HS2, but I don’t know the local reaction.

The second, minimal station boxes, may conflict with the aim of oversite development. Does Shoreditch High Street meet this target?

 

[eldomtom2]

The second, minimal station boxes, may conflict with the aim of oversite development. Does Shoreditch High Street meet this target?

In this context it is primarily talking about metro construction, so oversite development isn't really in the picture.

 

[Energy]

Kenilworth has a large cut and cover tunnel in an urban area for HS2, but I don’t know the local reaction.

Quite a lot of initial concern was about the woodland before the tunnel was extended, and some about a local small farm. Local concern now is particularly traffic, and hoping that it looks better once construction is done.

 

[HSTEd]

"Cheaper but more temporarily disruptive methods" often just offload externalities onto other parts of the state rather than actually save money. They should not just be used as an accounting trick to hide the cost of the project from the taxpayer.

For example, HS2 attempted to use an alignment on the surface along the New North Main Line which would have required the demolition and reconstruction of the Hangar Lane Gyratory.
It then transpired, once the costs of actually doing that, especially in disruption terms, were accounted for that it was cheaper just to dig a tunnel.

Cut and cover stations and tunnels might make sense in low density environments like most US cities, but they are not really practical in a UK setting.

How many obstacles do you think we can dig a 15+m wide trench through?
We don't have dual carriageways all over the place through all but the very densest urban areas.

 

[dosxuk]

Both lists mention it, but it really should be higher up - risk. At the moment the private sector is expected to carry the full risk of the works they are contracted to do - so they price it up based on the worst case option happening for every variable. It's endemic in the public sector - the desire for a single fixed price for a job outweighs the desire for getting good value for money.

The only people who still seem to believe that the idea of putting the risk on to the private sector means they'll do a better job because they'll have to pay for any problems are those who have spent their entire working lives in the public sector. In reality, the private sector wants to do things quickly and cost-effectively, because it's easier - but knowing things might go wrong just means they budget for that at the start - passing the cost along even when it isn't needed.

 

[Bald Rick]

Amongst many things missing in this analysis, two not covered are land costs and accounting treatment.

For land, the cost of land, ie the cost to a project of otsining the land and the rights to use it for a railway, are rather higher than in other countries. This is, in part, related to the south easrn quarter of the country being particularly developed and ’congested‘ - compare to Spian or France for example.

Accounting: Some jurisdictions count the cost before inflation, and some don’t count the tax paid by the project (eg the equivalent of VAT on services provided) as a project cost. I’m led to believe that some even take off the cost of income tax and equivalent paid to those working on the project, although I personaly think that is a dubious claim!

Then there’s the issues of what is counted as overheads, what is provided as ‘free issue’ (goods or services), the costs of time by management, etc etc.

It’s quite a complex subject, and getting hold of real world actual cost data at detailed level for international comparisons is notoriously difficult.

 

[stuu]

For land, the cost of land, ie the cost to a project of otsining the land and the rights to use it for a railway, are rather higher than in other countries. This is, in part, related to the south easrn quarter of the country being particularly developed and ’congested‘ - compare to Spian or France for example.

I have heard this argument when our seemingly ludicrous costs have been discussed, and I have to say I'm not convinced. A single KM of railway needs somewhere between 5-10 hectares of land, on average. Even if you are paying £2m a hectare that doesn't come close to explaining the disparity.

I do agree that a fair proportion of the difference is down to all manner of accounting differences though, it's not as construction firms here are making massively higher profits than elsewhere

 

[Bald Rick]

I have heard this argument when our seemingly ludicrous costs have been discussed, and I have to say I'm not convinced. A single KM of railway needs somewhere between 5-10 hectares of land, on average. Even if you are paying £2m a hectare that doesn't come close to explaining the disparity.

The cost of obtainimg the land is not the same as land costs.

 

[SynthD]

A single KM of railway needs somewhere between 5-10 hectares of land

Could you look at HS2 on satellite maps and make an estimate of hectare per km of their work zone?

 

[zwk500]

Several other threads relevant to this question:

Why are things so expensive?

https://www.express.co.uk/news/uk/1870214/new-30m-uk-train-station-open-closed-60-years So this is a single platform station with access road and car park for 40 cars. No footbridge, minimal infrastructure. Yet the cost is £30 million and it going to take 3 years.... Really? Surely a platform...

www.railforums.co.uk

https://www.railforums.co.uk/thread...ads-all-cost-more-to-build-in-britain.253578/ (this one also started by the OP)

Cost of new railway lines in 2024

What is the current cost of building a new railway in the UK? I'm thinking about the total cost, not just the civil engineering. There's a number of references I can find. Are there any better sources? East West Rail Central Section options Central section capital cost estimate (14.08.2019)...

www.railforums.co.uk

https://www.railforums.co.uk/thread...r-performing-net-zero-railway-by-2050.265949/ (this one also started by the OP).

 

[HSTEd]

I have heard this argument when our seemingly ludicrous costs have been discussed, and I have to say I'm not convinced. A single KM of railway needs somewhere between 5-10 hectares of land, on average. Even if you are paying £2m a hectare that doesn't come close to explaining the disparity.

I do agree that a fair proportion of the difference is down to all manner of accounting differences though, it's not as construction firms here are making massively higher profits than elsewhere

To a first approximation the of obtaining land (excluding land covered in things like apartment buildings) is the cost of the legal process of compulsarily purchasing it.
It will far exceed the actual value of the land.

EDIT: Edited to clarify language.

 

[eldomtom2]

Amongst many things missing in this analysis, two not covered are land costs and accounting treatment.

For land, the cost of land, ie the cost to a project of otsining the land and the rights to use it for a railway, are rather higher than in other countries. This is, in part, related to the south easrn quarter of the country being particularly developed and ’congested‘ - compare to Spian or France for example.

Accounting: Some jurisdictions count the cost before inflation, and some don’t count the tax paid by the project (eg the equivalent of VAT on services provided) as a project cost. I’m led to believe that some even take off the cost of income tax and equivalent paid to those working on the project, although I personaly think that is a dubious claim!

Then there’s the issues of what is counted as overheads, what is provided as ‘free issue’ (goods or services), the costs of time by management, etc etc.

It’s quite a complex subject, and getting hold of real world actual cost data at detailed level for international comparisons is notoriously difficult.

I had the chance to put your concerns to Alon Levy and received this in response:

We take into account land costs when we can, but they’re not even second-order elements. We likewise exclude financing charges whenever we can but include all other soft costs.

We’ve looked into some (far from all or even most) projects’ breakdowns to get a good sense of what matters and what doesn’t. For example, Italian projects are very good at breaking everything down, and if anything, the headline costs in Italy include more things than in the US (or UK), like rolling stock.

Land costs are always a minor portion. The big cost items are tunnels, stations, and systems. The US and UK have high soft costs, but that’s not land, that’s third-party design and engineering, which are less efficient than in Continental Europe (this can be seen from design and construction management contract costs – those don’t include land).

For any project big enough that its costs matter to the national or regional budget [we're confident all the expensive items are included in the project pricetags] – it’s just not possible to sweep a significant portion of the cost of such a project under the rug. Nor have we heard that smaller projects sweep e.g. utility relocations by budgeting them off-books.

 

[Bald Rick]

thanks.

I’ve no idea who Alon Levy is, but his comment that ‘land costs are always a minor portion’ rather misses the point. Asi stated, the cost of the land itself is often small except in develooped areas, where it is very high, and the UK has a lot of developed areas compared to most comaprtiive countries. It is the cost of the process for being permitted to obtain the land that is high.

 

[Manutd1999]

Regarding contracting, rail can maybe learn from companies who build infrastructure in more competitive industries, like wind/solar energy.

Multi-contracting is the norm in these projects to reduce overheads. Contracts aren't signed until the design is advanced and the scope is fixed. Aside from indexation against raw material prices and weather, the price is largely fixed as well.

The idea of giving a few large contractors a cost-plus contract to build a wind farm would be unthinkable, yet for HS2....

 

[ABB125]

Over the last year, I've been working (for an engineering consultancy) on a small (~£20 million) highway redevelopment scheme for a local authority, the main purpose of which is to reallocate space to cycle paths and pedestrian areas. The amount of work that I (and colleagues) do, which is then wasted because the client changes their mind, is unbelievable. Because of the way the scheme is structured, all official communication between client and designer (ie: the company I work for) has to go via the contractor who will be doing the actual work. This adds layers of time and cost to everything.

That's just a small example of why things in this country cost so much. There will be similar (bigger, worse) problems in the rail industry.

(As an aside, some of my colleagues are working on a seemingly never-ending project elsewhere which involves a small amount of work on Network Rail land outside a station (for clarity, as far as I'm aware it's all on the "safe" side of the boundary fence). The amount of paperwork (much of which is irrelevant but "has to be done anyway"), arguments with Network Rail frankly being unreasonable but ultimately getting the final say in what gets done, and general "can't do" attitude which gets mentioned is astonishing. Now imagine how bad it'll be when you're doing an actual rail scheme...)

 

[eldomtom2]

thanks.

I’ve no idea who Alon Levy is, but his comment that ‘land costs are always a minor portion’ rather misses the point. Asi stated, the cost of the land itself is often small except in develooped areas, where it is very high, and the UK has a lot of developed areas compared to most comaprtiive countries. It is the cost of the process for being permitted to obtain the land that is high.

I discussed this with Levy and got this in response:

Planning costs are included if they’re contracted out, or if they’re in the force account; if there’s a permanent civil service then they might not be, but the numbers I’ve been quoted in some low-cost environments are very low. In Spain, I was told that ADIF has permanent in-house staff for planning and that’s around 5% of project costs with some additional costs contracted out – and that’s for the entire planning process, not just the permission to build. It could just be that the UK makes that one step impossibly difficult. Then again, the UK also has laws favoring the state when it comes to safeguarding routes – in Italy safeguarding is for five years, not decades, and in the US it’s considered a taking so the agencies in theory have to pay for it and in practice don’t really do it, leading to private encroachments that sometimes scuttle projects (like the Downtown Extension in San Francisco).