Price and time saved difference between HS2 at 225mph and 185mph

[The Ham]

And the fact that capacity is generally given as the main justification for HS2 shows that it's been needed for a great many years.

Even if it wasn't needed before it soon will be, it's something which people often overlook.

Fairly low growth of 2.4% per year for 20 years means we would need to carry for 160 passengers in 2029 than there were in 2009, only even with COVID there are already (2024/25) 160 passengers travelling between London and the North West.

If you look back at my posts I've been saying that the 2.4% growth assumed in the HS2 modelling is a reasonable value (unlike that used in the HS1 modelling) and other than the free years of and after COVID it's often been growing faster than the model anticipated (and had again).

Yes we're still not at the 2019/20 year passenger numbers, but we're still ahead of the HS2 growth modelling by about 4 years (or about 20% earlier than expected). Anyone want to bet that we'll see zero growth between now and beyond 2029 so the HS2 growth model will be wrong?

That's not even needing HS2 to create lots of extra growth, that's growth we've already seen and it's baked in.

Whilst the London to the West Midlands is lower at 150 (that's not unexpected as it was aways said to be a more mature market), Scotland is much higher at 257 (yes, that's not a typo two hundred and fifty seven passengers for every one hundred in 2009).

If we roll forwards another 20 years we're not looking to carry 160 passengers but rather 260 passengers.

Even if the 390's were carrying 200 people on average (43% loading) in 2009, by 2049 that's potentially 538 passengers on average (so every train is more than 85% full on average - only ours nearly impossible to get an even distribution of people across the day).

Even with HS2 trains with 1,100 seats that's 49% loading on average which is higher than the loadings in 2009.

However, if growth hits 3% per year that's an extra 112 people per train (so over 100% loading on the 390's).

Although, that still doesn't account for any growth beyond that. Which HS2 does give us the potential to do.

It's easy to look at the here and now and say things are fine, but it doesn't account for what might happen.

 

[SynthD]

Secondly, it's not a reasonable comparison. HS2 is the first new long distance domestic route in the UK for well over 100 years, a flagship scheme bringing HSR to the UK for the first time (not counting HS1 as that was an international project).

LGV Bordeaux-Toulouse is a long way down the path of HSR development in France, whose flagship scheme (Paris-Lyon). 1981 wasn't it? Over FIFTY years ahead of HS2?

I’d use that as evidence if I wanted to argue that HS2 phase 1 is our starting point, where we learn how to get this type of project through Parliament and built by local people. Maybe France can guide us on what to do next, is it to ignore the non-constructive criticism and commission the next phases before the first has had a chance to prove itself?

 

[Austriantrain]

@Bald Rick ’s point is it’s not. LGV Bordeaux - Toulouse only started construction in 2024 and is due to open in 2032. They took 20 years to go from starting to shovels in the ground. (Initial studies 2002-2004). Whereas HS2 was half that time, announced by Adonis in 2009 and started in 2019.

True, but the reason was because that line is economically marginal. It would never even get planned in the UK.

 

[Bald Rick]

True, but the reason was because that line is economically marginal. It would never even get planned in the UK.

And yet we have EWR - admittedly not high speed, but a new railway nonetheless

 

[Palmerston]

Sorry, late to this.

I think I posted about that on the pther thread. But in short, if HS2 was already built to near Manchester (or even Crewe, or Lichfield), then you’ve already dealt with the difficult and thus expensive part. Therefore it is much,much chesper, snd better vslue for money, to continue to Scotland as an extension of that. Passenegrs from the big conurbations of W Mids and Manchester are on the route, services from Leeds can feed in via NPR.

Would it be fair to say the bit from Sheffield to Leeds is also expensive and difficult? How far north could you get before the Lake District and other bits of geography start to make a new line expensive? Would any bits north of Carlisle be relatively simple for a new stretch of line?

 

[redreni]

The OP's question - how much cheaper would it have been had a lower speed been specified and designed for at the outset - is surely of only academic curiosity? What the SpAds who briefed the Times on this need to explain, surely, is how on earth lowering the specs now is going to save an amount of money that makes the trade-off good value (or, indeed, save any money at all).

What I'm at least as interested in is what electricity costs do to the economics of operating a high speed railway. Obviously drag increases non-linearly with speed, and at the highest speed the extra energy needed to make the train go faster is quite significant. On the other hand, making the train slower increases staffing costs and at some point requires you to have longer trains and/or more trains to deliver the required capacity. Have projected energy costs changed enough, since HS2 was designed, to alter the balance of energy consumption vs operational efficiency in a way that would make slightly lower top speeds sensible?

 

[The Ham]

The OP's question - how much cheaper would it have been had a lower speed been specified and designed for at the outset - is surely of only academic curiosity? What the SpAds who briefed the Times on this need to explain, surely, is how on earth lowering the specs now is going to save an amount of money that makes the trade-off good value (or, indeed, save any money at all).

What I'm at least as interested in is what electricity costs do to the economics of operating a high speed railway. Obviously drag increases non-linearly with speed, and at the highest speed the extra energy needed to make the train go faster is quite significant. On the other hand, making the train slower increases staffing costs and at some point requires you to have longer trains and/or more trains to deliver the required capacity. Have projected energy costs changed enough, since HS2 was designed, to alter the balance of energy consumption vs operational efficiency in a way that would make slightly lower top speeds sensible?

It depends on how much slower.

TL;DR, the extra lease cost per return ticket London to Manchester is likely to be about 5p more (circa £2.65) by slowing the trains down and running then as 11 coach sets vs 16 coach sets, however the energy costs for the slower shorter train would be about £5.20 less by slowing the trains down and shortening them.

With the driver costs potentially being around a 20p saving for a return ticket by going faster

However, if the shower trains are full length the savings start to drop away quite quickly, with the extra lease costs rising by an extra £1.10 and the energy costs also rising by £2.25.

Key summary, yes there's a cost difference, but on a (say) £50 return ticket an circa £1.60 cost difference is hardly going to alter things by much.

Those complaining about high energy costs are often the very same people who don't want HS2, so focus on something which looks big but actually appears to have a fairly small impact on the overall running costs.

--------------------------------------------------------

The following gives the details of how the above were reached.

At 1:08 his London to Manchester with a 22 minute turn around you're already likely right at the limit of doing the return trip in 3 hours. With 3tph that's 9 sets.

That means of you slow it down at all it's going to be, at best, 9.33 sets (assuming you can rota the units with other services). Under phase 1 the extra isn't an issue, as you'd be running half length sets until the full link was to be built (originally only a few years).

However if you start to mess around with the assumed train lengths and you have to use 10 full sets you could see the following:

Original design under full opening with 16 coaches per set: 144 coaches but a capacity of 1,100 seats per train.

Phase 1 only with 11 coaches per set: 110 coaches but with a capacity of around 755 seats per train.

As such whilst you're not going to see an increase in rolling stock costs, the cost per seat would be higher.

Let's say that each coach costs a baseline of £40 per coach per hour that then gives you a value of:

Full HS2 - £5,760 shared between 6,600 seats, so a cost per seat of £0.87 per hour, so £2.61 for a return journey on a train 100% full.

Phase 1 - £4,400 shared between 4,530 sears so a cost per seat of £0.97 per hour, so £3.88 for a return journey on a train 100% full.

That's assuming £200,000 lease cost per coach per year, that they ruin 90% of the time running 15 hour days 363 days a year to get to that £40 per hour figure.

Obviously the loadings of the trains will adjust those numbers, as even with full capacity it's unlikely that you'd gain enough extra passengers to fill the 1,100 seats trains whilst not overflowing the 755 seat trains.

However if we assume 550 people per train that's going to be:
Full HS2, cost per ticket £2.61
Lower speed, shorter trains cost per ticket £2.76

However lengthen those slower trains to 16 coaches and the maths changes at you now need 160 coaches (vs the 110 for then previously or the 144 for the faster trains) and the cost rises to £3.87.

That's a long way to get to the point that the lease costs would be more per ticket, but not by very much.

The staff costs are more complex, however if we assume a charge rate of £110 per hour per driver (to allow for holiday, sick, training, hourly rate, test breaks, employment costs, etc.) then their costs would be:

£2 for a return ticket for the slower trains
£1.80 for a return ticket for the faster trains

Some might argue that's too high a charge rate just for a driver, but then I've not allowed for other staff time, so if anything that's likely to be a fair assessment. Obviously others could say that's too low, to which I'll say I didn't want to over do it and the overall outcome isn't significant anyway.

However, nor is the change in energy costs. However, by having fewer seats per train (as I understand it) there's actually proportional more drag per seat so you may also find that the energy savings aren't that big.

As I understand it, each coach you add are a bit of drag, but no where near the amount you have for the front of a train.

For example, whilst the total drag would be roughly double for the faster (225mpj) longer train compared to the slower train (185mph), the extra drag for the trailing coaches is broadly the same ratio, but only when traveling at the higher speed. When they are traveling at the same speed the amount of drag for the longer train would be higher but with the leading coach being 2.5 vs around 1 for a following coach the full set would be 17.5 vs the shorter set being 12.5 (at the same speed).

Therefore up to 185mph the cost difference would be limited, it's only as you go faster that you start to see the costs increase

Obviously the longer trains would have an extra energy cost due to carrying around extra coaches (roughly 40% more for like to like speeds), however this nearly doubles once the long train is running at 225mph vs the shorter train running at 285mph.

Assuming full speed end to end the power requirements for a single London to Manchester trip would be:

12-15kWh per passenger for the slower shorter trains

20-28kWh per passenger for the longer faster trains

At 20p per kWh the cost for the extra energy for the longer faster train would be £5.20.

However, that only applies all the time the slower trains are running as 11 coach trains, lengthen then to 16 coach trains and their energy use would go up by around 40% reducing that saving by £2.25.

Paired with the extra lease costs and the extra staff costs the net savings would fall to circa £1.60, which is hardly anything on a £50 ticket

 

[Austriantrain]

And yet we have EWR - admittedly not high speed, but a new railway nonetheless

Would you consider that economically as similarly marginal?

 

[takno]

Would you consider that economically as similarly marginal?

It's been pretty economically marginal so far

 

[Austriantrain]

It's been pretty economically marginal so far

Well, since nothing passenger-wise is running yet (at least, to my knowledge - has it changed?), that is not very surprising.

 

[Peter Mugridge]

Well, since nothing passenger-wise is running yet (at least, to my knowledge - has it changed?), that is not very surprising.

Just charters so far, most recently 800 032 running Oxford - Milton Keynes - Paddington on Friday last week ( 27th ) with a load of senior railway bods on board.

 

[Technologist]

125mph may have been too slow, how about 150mph or 175mph ?
I never could understand why they designed HS2 for 225mph in this relatively small country of ours....

Also need to be careful to avoid a "status quo" biase, current high speed trains take a long time to accelerate to 225mph. If we start designing battery assisted trains that are limited only by steel on steel traction and passenger comfort then we can get to 225mph in around 100 seconds and 5km. Battery electric cars are showing us that performance is basically a no cost option in a world where we manufacture millions of PM motors and TWh's of batteries every year. In which case the timesaving between stations 10's of miles apart is pretty close to the % differences in top speeds.

Future proofs it for what? I remember reading that SNCF doesn’t run trains over 320km/h because of power draw issues. There doesn’t seem to be any obvious reason why trains should go above 320-330, perhaps this is the realistic maximum that traditional trains can reach in service.

See above, a BEMU is not limited in peak power by the OHLE, it just needs enough power to supply the average power usage over the distance the line runs. Also SNCF trains do not represent the full capabilities of rail vehicle drag reduction.

If GBR is a proper research lead organisation it should be looking at the impacts of a Inductrack Maglev system similar to Nevomo was proposing. This would have the levitation elements laid alongside regular track so you could run "bi mode" trains that levitate at high speed and use regular rails at low speed. Regular rails being much better for switching, platforms and usage off the main track.

​

 

[NCT]

Back on planet earth, HS2 set out to buy a standard rolling stock product (or as close to it as possible) the market was offering / going to offer. We can debate the execution, but that was certainly the intent, and the right one.

 

[Technologist]

Here are a couple of examples off the top of my head:

At higher speeds you need to start looking at dynamic track-ground interaction - i.e. the mechanical waves in the ground forming a 'mach cone' ahead of the train. HS2 found this was particularly problematic for certain soil types which had low mechanical wave speeds. I don't know what solution was eventually decided but the talk was of having to excavate and replace several metres of ground under the formation.

Transitions from plain track alignment into structures becomes a problem. Say you have a deep cutting in clay, going into a tunnel. The ground in the cutting will heave upwards and this will continue for a long duration. But the tunnel won't move nearly as much. To maintain vertical track alignment you need to ensure the differential movement is limited. In normal circumstances you would use ballast to iron out the difference, tamping back into alignment. You can't do that with slab track so you need a permanent fix which usually comprises some sort of piled raft with progressively shorter piles going away from the tunnel. There would be a similar issue at the interface of high embankments with bridges/viaducts.

These are ground-related (betraying my specialism), doubtless there are more from other disciplines.

Almost sounds like we should have just put nearly all the track on a giant viaduct!

== Doublepost prevention - post automatically merged: 30 Mar 2026 ==

Back on planet earth, HS2 set out to buy a standard rolling stock product (or as close to it as possible) the market was offering / going to offer. We can debate the execution, but that was certainly the intent, and the right one.

Railways last hundreds of years, one of the reasons we needed HS2 was because all our railways were built with slow trains that don't like gradients. Future proofing the alignment should very much take into account more than the trains currently available on the market.

 

[Shrop]

Almost sounds like we should have just put nearly all the track on a giant viaduct!

It might be more than just coincidence that Chinese HSR has huge lengths of its construction on viaducts.

 

[The Planner]

Just charters so far, most recently 800 032 running Oxford - Milton Keynes - Paddington on Friday last week ( 27th ) with a load of senior railway bods on board.

Another one tomorrow.

 

[osmarks]

Almost sounds like we should have just put nearly all the track on a giant viaduct!

There was someone on Twitter arguing that the entire line should be in tunnels, to placate NIMBYs, have a more direct route and benefit from tunnelling economies of scale. It was quite a nice idea aside from (optimistically) roughly doubling the construction costs.

 

[quantinghome]

Also need to be careful to avoid a "status quo" biase, current high speed trains take a long time to accelerate to 225mph. If we start designing battery assisted trains that are limited only by steel on steel traction and passenger comfort then we can get to 225mph in around 100 seconds and 5km.

You'd need 1m/s^2 acceleration and about 30MW power output to do that; that's four times the maximum power of a TGV-M. Battery size required wouldn't be particularly extraordinary - less than 1000kWh storage would be enough, BUT it would need to discharge completely in 2 minutes. Then recharge as quickly. But by that point your problem would be more the speed restrictions due to track alignment when approaching built up areas which stations are inevitably located within.

 

[Technologist]

You'd need 1m/s^2 acceleration and about 30MW power output to do that; that's four times the maximum power of a TGV-M. Battery size required wouldn't be particularly extraordinary - less than 1000kWh storage would be enough, BUT it would need to discharge completely in 2 minutes. Then recharge as quickly. But by that point your problem would be more the speed restrictions due to track alignment when approaching built up areas which stations are inevitably located within.

I have done a decent amount of pre-lim design on that; I assumed a 1.3m/s acceleration limit 22MW peak discharge for the reference train base on a 5 car class 800. The battery was sized to equal the mass of diesels + fuel on the class 800 plus an allowance for using the battery as structure and thus deleting some of that structure and then putting the mass back into battery. This was assuming a 6C battery discharge which is pretty easy for an LFP type, the battery was 3.7MWh at end of life.

You are correct that HS2 wasn't quite built with this level of performance in mind the tunnels near London are not designed for 360kph running as it was assumed that trains will still be accelerating when they get to them. However were we to start building trains like this the whole network starts looking very different, you might actually hit 360kph then slow down, then accelerate straight back up again. We only need 2km and 22 seconds to go between 300kph and 360 kph. The whole UK rail network could start looking like a racetrack where we generate meaningful time savings by accelerating and decelerating on every short section of track.

We could also do a lot with Japanese style noses for tunnels and noise barriers further into cities, with high acceleration you'd also probably design HS2 quite a bit differently and put far more intermediate stations in and probably have all the trains stop at them. Essentially cover the UK in a national high speed metro, which would have a transformational impact on the country and economy.

 

[NCT]

You are correct that HS2 wasn't quite built with this level of performance in mind the tunnels near London are not designed for 360kph running as it was assumed that trains will still be accelerating when they get to them. However were we to start building trains like this the whole network starts looking very different, you might actually hit 360kph then slow down, then accelerate straight back up again. We only need 2km and 22 seconds to go between 300kph and 360 kph. The whole UK rail network could start looking like a racetrack where we generate meaningful time savings by accelerating and decelerating on every short section of track.

Purely speculating on my part, but I wonder whether you'd get a phone call from ASLEF.

 

[Technologist]

There was someone on Twitter arguing that the entire line should be in tunnels, to placate NIMBYs, have a more direct route and benefit from tunnelling economies of scale. It was quite a nice idea aside from (optimistically) roughly doubling the construction costs.

Another way of looking at it is that if you can halve the cost of tunnelling you can have high speed rail basically anywhere you want. I wonder what the cost of a development programme would be to increase the productivity of tunnelling by a factor of two, I suspect it would be an accounting error on the HS2 cost.

This is why we have HS2 fundamentally the wrong way around, the first port of call is to set out the top level network, the long term goal, then set up the body to build it, then start with a small project for them to learn all the lessons learnt so expensively on HS2.

== Doublepost prevention - post automatically merged: 30 Mar 2026 ==

Purely speculating on my part, but I wonder whether you'd get a phone call from ASLEF.

Forward them on to BRDC who will tell them to stop being jessies. Though after more thought they would probably oppose the idea primarily because any in-depth study would probably conclude that the weak link in any system which allowed trains to maximise speed by accelerating and decelerating frequently would probably be the driver!

 

[eldomtom2]

I have done a decent amount of pre-lim design on that; I assumed a 1.3m/s acceleration limit 22MW peak discharge for the reference train base on a 5 car class 800. The battery was sized to equal the mass of diesels + fuel on the class 800 plus an allowance for using the battery as structure and thus deleting some of that structure and then putting the mass back into battery. This was assuming a 6C battery discharge which is pretty easy for an LFP type, the battery was 3.7MWh at end of life.

You are correct that HS2 wasn't quite built with this level of performance in mind the tunnels near London are not designed for 360kph running as it was assumed that trains will still be accelerating when they get to them. However were we to start building trains like this the whole network starts looking very different, you might actually hit 360kph then slow down, then accelerate straight back up again. We only need 2km and 22 seconds to go between 300kph and 360 kph. The whole UK rail network could start looking like a racetrack where we generate meaningful time savings by accelerating and decelerating on every short section of track.

We could also do a lot with Japanese style noses for tunnels and noise barriers further into cities, with high acceleration you'd also probably design HS2 quite a bit differently and put far more intermediate stations in and probably have all the trains stop at them. Essentially cover the UK in a national high speed metro, which would have a transformational impact on the country and economy.

As best as I can tell the fastest accelerating Shinkansen only accelerate at 0.72 m/s². Considering how busy the Tokaido Shinkansen is, I think that if they could double that they probably would.

 

[melon68]

Another way of looking at it is that if you can halve the cost of tunnelling you can have high speed rail basically anywhere you want. I wonder what the cost of a development programme would be to increase the productivity of tunnelling by a factor of two, I suspect it would be an accounting error on the HS2 cost.

This is why we have HS2 fundamentally the wrong way around, the first port of call is to set out the top level network, the long term goal, then set up the body to build it, then start with a small project for them to learn all the lessons learnt so expensively on HS2.

== Doublepost prevention - post automatically merged: 30 Mar 2026 ==



Forward them on to BRDC who will tell them to stop being jessies. Though after more thought they would probably oppose the idea primarily because any in-depth study would probably conclude that the weak link in any system which allowed trains to maximise speed by accelerating and decelerating frequently would probably be the driver!

Tunneling where there are flooded mine workings would not be easy, these would probably need to be dewatered to below the level of the tracks. This would apply to the shallow coalfield between Leeds and Sheffield and between Manchester and Liverpool, amongst other areas,

 

[Bald Rick]

Would you consider that economically as similarly marginal?

Yep.

There was someone on Twitter arguing that the entire line should be in tunnels, to placate NIMBYs, have a more direct route and benefit from tunnelling economies of scale. It was quite a nice idea aside from (optimistically) roughly doubling the construction costs.

I well remember in the late 80s when the (then) Channel Tunnel Rail Link was being proposed to run through South London, a group of locals in Kent proposed an alternative that would be in Tunnel all the way from Folkestone to London. It was called RACHEL, although for the life of me I don’t know what it stood for!

 

[NCT]

It feels like after 4 pages of discussions we are still none the wiser. HS2 Ltd went with the basis that if you were going to build a railway the delta between 200km/h and 300/330/360/400km/h would be trivially low. There are commentators who are adamant speed is what's driven the runaway cost. Those commentators are never clear about which of the 200-300, 300-330, 330-360 or 360-400 they are referring to.

I suspect the truth, once it's eventually let out, is far more nuanced and somewhere in between. My suspicions are that, the 200-330 increments produces clear economic benefits that are well understood and captured in the business case and outweighs any corresponding incremental cost; the 330-360 increment has some benefit captured (360km/h running to recover from delays, which I believe is captured in the performance modelling) but it's possible this incremental benefit is smaller than the incremental cost; no benefits are captured in the 360-400 increment, and where instances of significant incremental cost were identified such increments could pragmatically have been value engineered out.

Is there anything published that shows the emerging / likely outturn unit costs? E.g. how much every single km in tunnel, cutting, or on embankment ('average' cutting and 'average' embankment) costs?

 

[osmarks]

Is there anything published that shows the emerging / likely outturn unit costs? E.g. how much every single km in tunnel, cutting, or on embankment ('average' cutting and 'average' embankment) costs?

We have estimated costs but I'm not aware of data on the actual ones: https://assets.publishing.service.g...1938c7e79a/hs2-cost-and-risk-model-report.pdf

I do enjoy how they have several sections on mitigating optimism bias.

 

[Sorcerer]

There was someone on Twitter arguing that the entire line should be in tunnels, to placate NIMBYs, have a more direct route and benefit from tunnelling economies of scale. It was quite a nice idea aside from (optimistically) roughly doubling the construction costs.

If there is anything HS2 has changed my stance on, it's NIMBYs. I can understand a degree of apprehension to a certain extent but for the most part they are so uncompromising that I almost don't want to even consider appeasing them anymore. People want more housing and economic growth but will oppose housebuilding and new infrastructure if it's too close to their homes.

In relation to reducing the max line speed I don't believe this will make the project that much more meaningfully cheaper. If a 200km/h line was going to be about 9-10% cheaper then it's going to be a much smaller saving bringing it down from 360 to 300. It's one of those short-sighted decisions that will cost us more in the long run, and as is typical with HS2, design changes will just add to the cost yet again.

 

[Technologist]

As best as I can tell the fastest accelerating Shinkansen only accelerate at 0.72 m/s². Considering how busy the Tokaido Shinkansen is, I think that if they could double that they probably would.

Those Shinkansen are close to the limits of current you can pull from a 25KV line.

The N700 pulls 17MW, the highest speed Chinese trains pull 20MW, Eurostar is 16MW.

Hence it can go at 0.72m/s2 until about 120kph.

With a battery a 16 car 400m long train could have ~65MW, which would mean it could go at 1.3m/s2 until it hit 260kph.

The 1.3m/s2 limit is traction and comfort limited, we would obviously be talking about every wheel powered. I’m pretty sure it wouldn’t take passengers too long to adapt behaviour to it, it’s not physically taxing to hold on while the underground accelerates.

My ideal network taking advantage of this level of power would have regular stops and have frequency more like a metro and be used as such.

 

[eldomtom2]

Those Shinkansen are close to the limits of current you can pull from a 25KV line.

The N700 pulls 17MW, the highest speed Chinese trains pull 20MW, Eurostar is 16MW.

Hence it can go at 0.72m/s2 until about 120kph.

With a battery a 16 car 400m long train could have ~65MW, which would mean it could go at 1.3m/s2 until it hit 260kph.

The 1.3m/s2 limit is traction and comfort limited, we would obviously be talking about every wheel powered. I’m pretty sure it wouldn’t take passengers too long to adapt behaviour to it, it’s not physically taxing to hold on while the underground accelerates.

My ideal network taking advantage of this level of power would have regular stops and have frequency more like a metro and be used as such.

My point was rather that if using BEMUs would unlock much greater levels of acceleration, why haven't we seen any railway do so?

 

[The Ham]

My ideal network taking advantage of this level of power would have regular stops and have frequency more like a metro and be used as such.

Where are there places which would justify 3tph (or better) where such an acceleration rate would have a material impact on journey times?

For most people they would feel like their train is faster by not having to wait as long between services.

Most people want at least 2tph, however factor in some time for delays (it's not uncommon for a commuter to get the train that gets then to work at 08:29 even though they could get there at 08:59) especially if they have to change.

Typically passenger modeling assumes this delay to be half the gap in service provision.

That means that the "saving" from having a train service running at 3tph vs 2tph is 5 minutes. Even at maximum performance (1.3m/s2 from 0-360kph at each stop) that's not going to save as much as the more frequent service, unless there's more than 4 station stops.

However, most trains with more stops (i.e. typically those getting up to 100mph) are going to have higher acceleration rates (even the third rail 444's can do 1m/s2) which reduces the acceleration advantage.

As such that's a lot of engineering for not a lot of savings (basically none under normal conditions, as you shouldn't be timetabling for such tight margins).