HS2 Reset: Evidence to Transport Committee - 9 July 2025 "Total overspend on high-speed rail could be as much as 100%"

[Peter Mugridge]

Are any trains for overseas made here?

Yes, most recently some monorails from Derby to, I think it was Egypt?

 

[class26]

Yes, most recently some monorails from Derby to, I think it was Egypt?

Yes, Cairo metro is the latest I believe

 

[JW4]

Swinford article in the Times says it is expected to be announced on Tuesday that it will be delayed and reduced in speed

Latest news & breaking headlines

The latest breaking UK, US, world, business and sport news from The Times and The Sunday Times. Go beyond today's headlines with in-depth analysis and comment.

www.thetimes.com

On Tuesday Heidi Alexander, the transport secretary, is expected to confirm that the speed for the line will be reduced to 320km/h to reduce costs, while delivery of the project will be further delayed from previous plans to get trains running by 2033. The first trains between London and Birmingham are expected to run from 2035 at the earliest.

 

[FMerrymon]

Swinford article in the Times says it is expected to be announced on Tuesday that it will be delayed and reduced in speed

Latest news & breaking headlines

The latest breaking UK, US, world, business and sport news from The Times and The Sunday Times. Go beyond today's headlines with in-depth analysis and comment.

www.thetimes.com

Whilst the overall story might be correct, The Times has serious issues with accuracy...

Latest estimates suggest that this line alone could cost as much as £100 billion, the equivalent of £1 billion a mile

Its 140 miles, nowhere near 1bn a mile.

In 2011 the price tag for the scheme was £32.7 billion for a line that would have connected London, Manchester and Leeds

No it wasnt.

== Doublepost prevention - post automatically merged: 17 May 2026 ==

In any case, for the reduced frequency, it's not going to need 360, and 320 will not be massively slower than the 330 the journey times we're calculated at.

 

[Nicholas Lewis]

Swinford article in the Times says it is expected to be announced on Tuesday that it will be delayed and reduced in speed

Latest news & breaking headlines

The latest breaking UK, US, world, business and sport news from The Times and The Sunday Times. Go beyond today's headlines with in-depth analysis and comment.

www.thetimes.com

Something is seriously wrong if its going to take till 2035 so looking forward to seeing what the reasons are

 

[JonathanH]

Something is seriously wrong if its going to take till 2035 so looking forward to seeing what the reasons are

Isn't it just another spreading of costs, because of cost inflation, and providing sufficient time for testing? Does slowing the maximum speed down to save construction money also cost design time and money?

If there is an annual budget for HS2 costs, and costs inflate, the only way to square the circle is to extend the construction period. It all points to reasons why it is silly to talk in the King's Speech about construction in Manchester or speculate about what Andy Burnham might do to ensure HS2 will reach Manchester. The reality seems to be that the primary control on progress remains the finite annual funding available.

 

[fishwomp]

Its 140 miles, nowhere near 1bn a mile.

£715M a mile is so much more reasonable!

[on 32bn to Manchester 2011 pricing]

No it wasnt.

Well.. to be honest I'm not sure but let's see what Justine Greening announced in 2012: https://www.gov.uk/government/speeches/making-hs2-happen

Some commentators have said we are spending £32 billion building a new line between London and Birmingham.

Well I have good news for them. £32 billion will buy us a national high speed network, linking London and Birmingham with Leeds, Manchester, South Yorkshire and the East Midlands - and forming the base of what we hope is an even bigger network extending to Scotland in the future.

 

[FMerrymon]

£715M a mile is so much more reasonable!

Nominal or constant prices? If constant, what years prices? 100bn not confirmed anyway, but even if so that's still considerably less.

Well.. to be honest I'm not sure but let's see what Justine Greening announced in 2012: https://www.gov.uk/government/speeches/making-hs2-happen

She announced a mean estimate, without rolling stock, of 32.7bn in Q2 2011 prices. That last bit is essential.

 

[Nottingham59]

From the Times article:

"In 2011 the price tag for the scheme was £32.7 billion for a line that would have connected London, Manchester and Leeds and was due to open by this year."

No it wasnt.

Yes is was, as you yourself say:

She announced a mean estimate, without rolling stock, of 32.7bn in Q2 2011 prices. That last bit is essential.

 

[eldomtom2]

According to the Bank of England's inflation calculator, £32.7bn in 2011 prices is £49.3bn in 2026 prices.

 

[FMerrymon]

From the Times article:




Yes is was, as you yourself say:

No it wasn't

She announced a mean estimate, without rolling stock, of 32.7bn in Q2 2011 prices. That last bit is essential

 

[zwk500]

Does slowing the maximum speed down to save construction money also cost design time and money?

Not meaningfully, I'd imagine, given that the big civils are all now designed and in construction.

 

[Snow1964]

public Accounts committee going to look at this, and a Parliamentary session on Wednesday too.

18th May 2026
Two Select Committees today announce the next steps in their scrutiny of ongoing progress in delivering the troubled High Speed 2 (HS2) railway.

• Watch: 'The project threatening to cost taxpayers £100bn'

The Public Accounts Committee (PAC) has launched an inquiry scrutinising delivery of HS2 and the Euston station, while the Transport Committee will hold an evidence session with Rail Minister and the chief executive of HS2 this week.
Ahead of this Committee activity, PAC Chair Sir Geoffrey Clifton-Brown and Transport Committee Chair Ruth Cadbury met to discuss unanswered questions and areas of scrutiny for both their Committees around the historically troubled project. You can watch the video of their conversation linked above.
The Chairs’ discussion followed a roundtable held with the Department for Transport (DfT), HS2 Ltd and other stakeholders in the Euston project which informed the questions their Committees now have for those responsible for HS2’s delivery. You can read the lettercontaining the main themes that arose from those discussions.
It is now more than 14 years since HS2 was given the go-ahead by Government, with costs escalating throughout the lifetime of the project and the entry into service date slipping.
In an interim assessment in March 2025, HS2 Ltd chief executive Mark Wild concluded that it would not be possible to deliver HS2 in 2029-2033 and said there was a need for a reset to deliver the project “at the lowest feasible cost”.
The reset was initially expected to be published in Spring this year but has since been delayed, with the Secretary of State yet to indicate when it will be complete.
Wild will take part in the Transport Committee’s evidence session at 9.15 on Wednesday 20 May, alongside Rail Minister Lord Hendy and senior DfT official Dean Creamer.
MPs are likely to ask when the full reset will be complete, and to press witnesses on the expected entry into service date for HS2 and potential plans to reduce the speed of trains.
The PAC inquiry will kick off with an evidence session to be held before the House of Commons rises for its summer recess on 17 July. Likely topics for the inquiry will include the funding, costs and timeline of both Euston and the wider project, as well as the community impacts and benefits of each.

https://committees.parliament.uk/committee/153/transport-committee/news/213577/spotlight-on-hs2-commons-select-committees-scrutinise-projects-progress/

 

[Technologist]

If we abandon the fiction of meaningful competition in UK rolling stock manufacture, we probably can.

The UK (EDIT: National Rail) passenger rolling stock market is probably 600 or so vehicles per year. It could sustain a factory and modern technology would allow us to fill all the roles we require with only two or three designs of train.

Right now we have way too many factories because politicians like cutting ribbons, and rolling stock orders are distributed based on the necessity to avoid embarassing plant closures rather than quality.

Agree, industries tend to go through stages or vertical and then horizontal integration. Horizontal integration is best for reducing costs in a mature industry vertical integration is where you go when you need to change things. Tesla and it's Chinese clones are vertical integrators, the transition to BEVs was not a "deep tech" revolution where a single radical technology came in an changed everything. It was more a case of integrating a bunch of advanced technologies that had been wholly or partially demonstrated in other industries.

At the moment we cannot innovate because the people setting the strategy have no design capability and the people who win the train orders are orientated around compliance with very complex contracts, they also outsource a lot of their technology so struggle to innovate rapidly. If you want innovation you need a scenario where you can actually build something and make changes as you go and find things out.

It would be very expensive to build a "super BEMU" prototype using BEV technology via the current processes and supply chains, if you did it in house or by a start-up you could probably do it for less than £100m.

There are also some other models that you could use for rolling stock developments. During the cold war in the case of land armaments frequently what would happen is that a state agency like Fighting Vehicles Research and Development Establishment (FVRDE) who would build the prototypes and then industry would industrialise it. A lot of the kit the British army still uses was designed that way.

It's technically feasible to do a lot of things, or it will be as the future is a really long time. But when or if it becomes economic... and that's a critical uncertainty that is very hard to foresee and a reason why futurology is very hard. The downsizing of Euston would therefore be a massive error because you can imagine rail systems in 100 years that combines virtual couplings and advanced signalling to reduce headways, but the people side is harder to transform as the logistics of getting people on and off trains hits hard limits around terminal capacity, walking speeds and so on which requires space.

Japan manages 17 tph on 6 platforms, people and their ability to move are not untradable values. Psychological solutions tend to be the cheapest and most rapidly deployed ones but they are something that engineers don't like working with because they are uncertain and not really in their wheelhouse.

You could probably get people off a train faster by giving the first 1/3 of people to tap out some sort of reward and by targeting support to slow movers. Probably stuff could be done with luggage in terms charges or even moving to a concept of "hold baggage" where items over a certain size are not stored in the cabin. Trains could definitively be turned around faster, an F1 team can change tyres in less than 2 seconds vs 4-5 seconds in the 90's with a combination of optimisation, practice and technology, watching trains at St Pancras they appear to mostly sit around doing nothing.

More broadly, stuff like virtual coupling is a technology that isn't particularly difficult, its not deep tech its an integration problem. What GBR needs to do is start finding places to try stuff out where it isn't massive stakes. This would need to be a change in attitude as these technology trials would under no circumstances pay for themselves, they would make sense when applied more broadly.

An Intercity service where they brake as hard as tube trains. So people are sitting until it’s stopped, pick up luggage, exit, new people come on, put luggage somewhere secure enough for those acceleration rates, sit down, within 90 seconds? Super speedy trains with super performance service timings need super humans to always do the right thing for that plan.

People stand up and walk to the doors on tube trains and on buses that brake considerably harder than a tube train. My view is that this train would be configured like a metro with double doors opening both sides so people can get on and off at the same time. It would be coming at metro frequencies which will spread the load out and make it not too much of hardship if you can't board the train, you just get on the next one. Ultimately people would match their behaviour to the mode of transport, people would bring luggage on the basis of 90 second dwell time, people manage to go to Heathrow on the underground. The high speed metro would be a different user experience than current long distance rail, people can go years without getting on a train, but how many people just make one tube journey a year?

I think the real transformation will be super train on regular lines, with that level of acceleration they could beat current express end to end times and stop at all the stations plus turn any decent straight into an opportunity to crack 200mph. The use of an eddy current brakes means that they could quite easily decelerate into corners at 1.3ms-2, and still be able to get around the bends even if they miss their braking points by quite a margin, again we'd be talking about decelerations well within what we commonly deal with on busses where people also stand.

 

[NCT]

You could probably get people off a train faster by giving the first 1/3 of people to tap out some sort of reward

Good luck getting that past Health and Safety.

 

[SynthD]

It would be very expensive to build a "super BEMU" prototype using BEV technology via the current processes and supply chains, if you did it in house or by a start-up you could probably do it for less than £100m.

I agree with the general point, but surely if the current process is so burdensome for a contract, it would be for internal commissions too. It's a child of the burden from above to contract out the work rather than do it in house. Once any part of that is fixed, an existing supplier will be competitive and quicker for now.

Japan manages 17 tph on 6 platforms, people and their ability to move are not untradable values.

Huh? Is their ability a lesson we can learn from and achieve?

My view is that this train would be configured like a metro with double doors opening both sides so people can get on and off at the same time. It would be coming at metro frequencies which will spread the load out and make it not too much of hardship if you can't board the train, you just get on the next one. Ultimately people would match their behaviour to the mode of transport, people would bring luggage on the basis of 90 second dwell time, people manage to go to Heathrow on the underground. The high speed metro would be a different user experience than current long distance rail, people can go years without getting on a train, but how many people just make one tube journey a year?

It's not the journey people are expecting to take, so they are either scared off into cars or take a few yearly journeys to learn what they 'should know'. I would focus on the positive and inducing parts of the customer experience, eg waiting rooms under the full length of the platform, or the non-customer parts, eg looser scheduling for drivers and cleaners so they're waiting. NIMBYs will be even worse if everyone with a walking stick is de jure excluded from your high speed line.

 

[FMerrymon]

public Accounts committee going to look at this, and a Parliamentary session on Wednesday too.

https://committees.parliament.uk/committee/153/transport-committee/news/213577/spotlight-on-hs2-commons-select-committees-scrutinise-projects-progress/

PAC chair thinks freight should run on HS2? Someone want to explain to him what it's all about?

https://youtu.be/LbhynaoFjEo

 

[The Ham]

It would be very expensive to build a "super BEMU" prototype using BEV technology via the current processes and supply chains, if you did it in house or by a start-up you could probably do it for less than £100m.

The class 800's could fairly easily be changed to a BEMU very easily by switching out the diesel packs with batteries.

Any new build wouldn't be complex to deliver given that there's been various trails with batteries in the UK with retrofited packs to existing units.

 

[Nicholas Lewis]

Isn't it just another spreading of costs, because of cost inflation, and providing sufficient time for testing? Does slowing the maximum speed down to save construction money also cost design time and money?

The alignment is set the size of the tunnels established for 360kmh so the big win civil options have been and gone. Cant see how the signalling will be anymore cost effective. They've ordered the grid connections based on the demand the full fat HS2 was going to need and you have to pay Nat Grid up front for that so hows that going to save money now. Maybe the catenary design can be down specced slightly. To my mind the only big option left maybe is to revert back to ballasted track in some areas and maybe the trains can revert back to a proven 320kmh design. Mind you as your rightly say any redesign just adds back in costs.

If there is an annual budget for HS2 costs, and costs inflate, the only way to square the circle is to extend the construction period. It all points to reasons why it is silly to talk in the King's Speech about construction in Manchester or speculate about what Andy Burnham might do to ensure HS2 will reach Manchester. The reality seems to be that the primary control on progress remains the finite annual funding available.

They do have a annual budge of c7B but it gets indexed linked each year although only for next 3yrs. Anyhow id this is to save money it should be freeing up budget to accelerate works.

 

[LNW-GW Joint]

The alignment is set the size of the tunnels established for 360kmh so the big win civil options have been and gone. Cant see how the signalling will be anymore cost effective. They've ordered the grid connections based on the demand the full fat HS2 was going to need and you have to pay Nat Grid up front for that so hows that going to save money now. Maybe the catenary design can be down specced slightly. To my mind the only big option left maybe is to revert back to ballasted track in some areas and maybe the trains can revert back to a proven 320kmh design. Mind you as your rightly say any redesign just adds back in costs.

The orders for the rail systems (slab track, rail, OHL, signalling etc) were only placed in, I think December, after a long review, so I doubt they can be changed significantly.
The rolling stock contract is not finalised completely yet, but changes to the 54x8-car spec are likely to cost more, not less, unless the number of trains is reduced.

 

[HSTEd]

The alignment is set the size of the tunnels established for 360kmh so the big win civil options have been and gone. Cant see how the signalling will be anymore cost effective. They've ordered the grid connections based on the demand the full fat HS2 was going to need and you have to pay Nat Grid up front for that so hows that going to save money now. Maybe the catenary design can be down specced slightly. To my mind the only big option left maybe is to revert back to ballasted track in some areas and maybe the trains can revert back to a proven 320kmh design. Mind you as your rightly say any redesign just adds back in costs.

Quality control requirements for 360km/h are likely to be substantially more stringent than for 320km/h.

Tunnel size is probably not a significant cost driver in.any case.

 

[Snow1964]

The rolling stock contract is not finalised completely yet, but changes to the 54x8-car spec are likely to cost more, not less, unless the number of trains is reduced.

It's very unlikely the train builders will accept a reduction in the rolling stock contract value. If they do agree to a variation, bound to be something that increases contract value and their profitability.

There is an old adage that variations never save money in long run

 

[Peter Sarf]

.....................

I think the real transformation will be super train on regular lines, with that level of acceleration they could beat current express end to end times and stop at all the stations plus turn any decent straight into an opportunity to crack 200mph. The use of an eddy current brakes means that they could quite easily decelerate into corners at 1.3ms-2, and still be able to get around the bends even if they miss their braking points by quite a margin, again we'd be talking about decelerations well within what we commonly deal with on busses where people also stand.

Well as someone who was injured by the aggressive acceleration of an electric bus I have to say I think buses need to slow down.
I won't be travelling on fairground ride like public transport unless we are all seated and strapped in before the movement starts.

 

[HSTEd]

Well as someone who was injured by the aggressive acceleration of an electric bus I have to say I think buses need to slow down.
I won't be travelling on fairground ride like public transport unless we are all seated and strapped in before the movement starts.

Studies (open access academic journal article on the topic with graphs) have shown that accelerations and decellerations of around 1.1-1.3m/s/s are tolerated well by the vast majority of the population, assuming they are at low values of 'jerk' (ie. the acceleration comes on or off smoothly over a few seconds). Metro lines across the world do these accelerations all day every day, with the Victoria Line regularly achieving 1.3m/s/s.

With strap hangars and grab bars I don't see much problem in making trains do those sorts of accelerations - which would allow a train to go from 0 to 200km/h in ~45 seconds or so. The challenge is in getting rolling stock that can do that all the way to high speeds.

EDIT:

I've checked, and the journal in question permits reproduction of images with credit - so credit to the authors of the article and the Urban Rail Transit journal.

 

[FMerrymon]

They do have a annual budge of c7B but it gets indexed linked each year although only for next 3yrs. Anyhow id this is to save money it should be freeing up budget to accelerate works.

The reschedule in March 2023 is thought to have added £3.1bn in 2019 prices to costs of phase 1 according to the NAO. I do think this annual limit ends up driving further costs.

Cant see how the signalling will be anymore cost effective

ATO is going apparently. This would have been needed for the 18 trains an hour spec and running up to 360kmh. It's perhaps the bit with the highest risk for testing.

 

[InTheEastMids]

Japan manages 17 tph on 6 platforms

I think the issue is translating that to a UK context. Projects like HS2 require coalitions of support to move forward, so at the Concept phase, HS2 would need a CEO capable of winning an argument with Government (and the electorate) along the lines of "Yes it'll work... provided we can persuade British people to act like Japanese people" and successfully enduring Daily Mail photo mock-ups of HS2 with Japan-style pusher people.

Then it actually needs to work in operation - i.e. avoiding public anger about people over-carried/left behind or unceremoniously bundled on and off trains by crowds. Or the network getting logjammed if trains can't hit dwell targets, and then service reductions until it's all sorted out. HS2 has (with hindsight) showed itself not to be able to manage all the normal risks around scope/cost/schedule control and the latest criticism relates to HS2 having been excessively ambitious. So I don't think loading it up with more ambition and risk would have been the way to go.

However, I do agree that society evolves over time, it may look more like your vision in 50-100 years time, provided the network does get extended to be a basically segregated network connecting London to all the major cities of the Midlands and the North.

It's not the journey people are expecting to take,

Agree with this, the customer expectation is today's European inter-city travel environment, the obvious comparators are Cl390 and 800 in the UK, and modern TGV/ICE. I find imagining myself as a tabloid editor (it's not nice but it is helpful) and again the media reaction if Europe has these nice double-deck inter city trains and HS2 is seen to be giving us Thameslink stock with very pointy ends.

stuff like virtual coupling is a technology that isn't particularly difficult, its not deep tech its an integration problem.

However the Elizabeth Line shows how hard technology integration can be, and not seeing evidence that a repeat performance is very unlikely. Perhaps GBR will be better able and more focused on using test tracks for this kind of thing.

The class 800's could fairly easily be changed to a BEMU

Class 769 says "Hi" here - I'm not saying that batterification doomed to fail, but it's naïve to assume that a significant reengineering of a train in mid-life is a cake-walk. There was the TPE demo but that wasn't in regular/intensive passenger service, so is a trial, not a pilot.

Any new build wouldn't be complex to deliver

Classes 701 and 810 say "Hi" (although I don't think in either case it's the traction systems)

tolerated well by the vast majority of the population

Interesting thanks. Firstly I wondered what the acceptable level of jerkiness and acceleration is for a transport project as we move to a world where it is the human, rather than the technology that is the limiting factors - in a world that's designed for inclusivity and more older people.

Secondly, thanks for introducing me to the unit of Jerk, and the potential for a new geeky insult. "Piers Morgan is absolutely 1,000,000 m/s/s/s".
People that work on this topic are probably bored of it, but it was new to me.

 

[Peter Sarf]

Studies (open access academic journal article on the topic with graphs) have shown that accelerations and decellerations of around 1.1-1.3m/s/s are tolerated well by the vast majority of the population, assuming they are at low values of 'jerk' (ie. the acceleration comes on or off smoothly over a few seconds). Metro lines across the world do these accelerations all day every day, with the Victoria Line regularly achieving 1.3m/s/s.

With strap hangars and grab bars I don't see much problem in making trains do those sorts of accelerations - which would allow a train to go from 0 to 200km/h in ~45 seconds or so. The challenge is in getting rolling stock that can do that all the way to high speeds.

EDIT:

I've checked, and the journal in question permits reproduction of images with credit - so credit to the authors of the article and the Urban Rail Transit journal.

Well it was a violent jerk in my case, just got on and half way up the stairs. It was unexpected. But to expect every one to have got on, found a seat or somewhere to hold on before getting moving with high rates of acceleration is unrealistic. Be it bus or train. Not everyone is an agile teenager.

 

[The Ham]

Class 769 says "Hi" here - I'm not saying that batterification doomed to fail, but it's naïve to assume that a significant reengineering of a train in mid-life is a cake-walk. There was the TPE demo but that wasn't in regular/intensive passenger service, so is a trial, not a pilot.

I would suggest (as the trial didn't appear to suggest significant issues) that taking out a diesel engine designed to provide electricity and be removable and replacing with a battery (of the same weight) would be easier than adding diesel engines to a EMU which was never designed to take the extra loadings.

Classes 701 and 810 say "Hi" (although I don't think in either case it's the traction systems)

There's always a risk that things will go wrong with any project, I'd that stopped us from doing anything we'd still have slam door stock.

 

[ScottGallagher]

Deleted

 

[JonathanH]

So by the sound of it and watching Heidi Alexander live in the commons HS2 will depart from Euston to Handsacre Jn as late as 2043. Wow.

At pains to also make the point that abandoning HS2 now would cost as much as finishing it given remedial costs involved.