HS2 line speed/performance profile example

The value proposition for HS2 is a lot simpler than the press typically makes out – it’s actually summarised in the speed profile. Speed profile? A map of changes in speed as trains go through or stop at stations and signals – all railways have them.  Why so important?  Trains don’t go from stop to full speed very quickly, especially “high speed” (5-10 mins), and once you work through the route to create a profile, you can work out capacity, where the pinchpoints are, and even how much power you need, section by section.   And of course how long a train runs at maximum line speed.

For HS2, you can safely assume that the reason it tears though Middle England without stopping is because the speed profile would look godawful if it did stop – it would rarely get up to the mythical “high speed”, and then only for a few minutes at a time.  If the speed profile looked bad, there’d be no point in it being built as “high speed” and substantial cost and some valuable ancient woodland could be saved through more curved lines, and it could stop and serve the passengers at intermediate points, which would make those affected much happier.  The project would then begin to look like a much more democratic enterprise, a lot of the rail-induced, testosterone-fuelled excitement and criticism would simultaneously ebb away.

The truth is that for the most part, the UK isn’t really a large enough country to warrant true ‘high speed rail’… The 30 minute saving London-Birmingham would all but disappear if a couple of stops were added.

The truth is that for the most part, the UK isn’t really a large enough country to warrant true “high speed rail”, because most of the intercity journeys where the capacity is needed are at most around 100 miles long, unlike say, Paris-Lyon (287 miles), Paris-Marseille (482 miles) or Madrid-Barcelona (386 miles).  Tellingly the nonstop Barcelona service is 2h30, but 3h10 when stopping at a handful of stations.  Equivalently, the 30 minute saving London-Birmingham on HS2 would all but disappear if a couple of stops were added, which is why there are none.  However, at 49 mins, this journey is below accepted ‘high speed’ viability of 1 hour.  The 200 miles London – Manchester is perhaps where the service approaches viability – as long as it’s a regular enough nonstop service. 500 km/300m is an optimum distance for the such lines.

This is the bind that UK high speed rail finds itself in, and why it’s neither the most cost-effective nor most appropriate solution in most cases.  The one stretch actually best suited to high speed rail is the 393 mile long and fairly straight East Coast mainline, with its fewer well-spaced natural stops London – York – Newcastle – Edinburgh, a possible stop around Peterborough, connecting the capital cities, and with its many longstanding overhead line problems needing properly fixing, but there are only make-do-and-mend plans for it !  It is absolutely no coincidence that early steam speed records for the Flying Scotsman and Mallard were set on this line. 

At the time those steam records were being set, Japan had already conceived the high speed (160mph) Shinkansen train Tokyo-Osaka, 515km long, coming to fruition nearly 60 years ago.  If the UK were trying to make a World-beating high speed rail statement costing a significant proportion of GDP, it would look like a proper 300mph+ straight high speed line London – Edinburgh.

Crossrail is finally beginning to realise that it should have built a representative live test track for systems integration test, as was done for a rail project your scribe worked on in Singapore, as opposed to a static “Integration Test Facility” – a shed where mockups of systems and simulators are tested together.  The test track removed a big dependency, and much reduced the need for the many months, err.. years, of “trial running” which Crossrail still requires, and which can be invalidated if anyone makes significant changes, so conventionally comes right at the end of the project. That puts it, in bright red, squarely on the “critical path”, as a massive risk, as the Mayor has discovered.

A quick perusal of the latest Crossrail signalling and testing fact sheet (Nov 2019) provides the gory details of the different generations of incompatible systems they are trying to integrate. 

Despite it appearing to be “just a technical problem”, if a seasoned tech-savvy leader with hands-on integration experience had been in place at the highest level early in the project, he would have insisted on either building a proper test track to take it off the critical path, or a totally different approach to remove the risk in the first place.

Way above and beyond?  Yes, the test track was built in Northern France (adjacent to the train factory), and somewhat reluctant Singaporean and Japanese staff were dispatched there the live in a hotel for a year or so to test and supervise.  Result? Project completed on time.  The delay incurred before opening, was not because the trains didn’t work, but because the inexperienced operator (previously only running buses) was nervous and didn’t want any bad PR.  Bad PR in Singapore is a train delayed by 10 minutes.

The operator shouldn’t really have been nervous, the trains are, after all, fully automated and driverless, launched through the automated train wash into service at the click of a mouse.  Is this some recent invention or fanciful future concept?  No, the World’s first automatic, fully underground heavy railway has been in reliable operation for 18 years, and recently achieved 4 million train-km between delays!  Watch out TfL –  Tripadvisor rating 96% excellent/very good.

A lack of understanding of the somewhat opposing objectives of the project “sponsors” in big project parlance (read government) and those building it.   Sponsors – mythical chequebook in hand – don’t appear to understand the game of megaproject consultancy, despite there often being a revolving door between the two in the higher echelons. 

In a laudable enough aim of reducing risk by employing well-known consultants, sponsors fall into the trap: too scared to either manage the risks themselves or deal with the fall-out should all not go to plan, they pay others to take the risk.  But make no mistake, the sponsor still takes a big risk, regardless.   It’s almost always a fallacy that (in the parlance) “the risk is transferred” to X or Y party because they are contracted to design/build this bit or that, because a) the contractor priced-in the risk, and added a margin – government is paying, in a kind-of “heads they win, tails you lose” scenario, b) by chopping the project up into hundreds of “work packages” amongst multiple contractors, the lines of responsibility are muddied, the opportunity for blame-shifting is substantial, and c) the important risks are often not what’s on the project risk register anyway, and all that money didn’t really cover those risks, such as huge delays for which there’s a whole subculture of expertise in excuse-making, the latest of which will be COVID-19.   In fact the pandemic will be manna from heaven for this type of enterprise as it can be termed unforeseeable Act of God (“FM” – Force Majeure), which will trigger all sorts of clauses, insurances etc, as it has already at Crossrail.  This, despite a pandemic being no.1 on the government’s own risk register for decades – case closed! 

The political fallout will be as bad for those involved as if they’d built it for a fraction of the price but had to rebuild a section or two because they didn’t work so well, as the Mayor is about to find out.

West London College (RIBA picture library)

Not that long ago, a local government Architects’ Department would have designed major civil infrastructure, and perhaps helped manage the construction, certainly for buildings such as West London College.  Further back in history, Sir Joseph Bazalegette, much celebrated for Victorian sewers and our long-suffering bridge, wasn’t a staffer at a top consultancy, or some other outsider, he was Chief Engineer of London’s Metropolitan Board of Works.  The clue is in the name –  Chief ENGINEER of the Metropolitan Board of Works – a public body, for which he engineered, not politicked – he personally signed off most of the designs, as records testify.

The game of megaproject consultancy has long been “bums on seats”, not necessarily particularly experienced ones, but ones that are covered by that large risk pot and the name of the consultancy. Your scribe has sat in an ad-hoc meeting with a young draughtsman from a name consultancy, next to a World expert in architectural lighting (whose name it transpired is the company name), and many others in-between.  Suffice to say that the trick was to work out who was who, before putting foot-in-mouth, as it was not always immediately obvious if the intros were skipped.  We digress. The trick of the consultancy is to a) charge out the young draughtsman at a substantial rate with b) as many others as can be mustered. In fact there is a rather unedifying scene at the start of many a project with consultancies jockeying to bag as many desks as possible for themselves, whether or not they can name the individuals who might sit at them – they are the meal tickets. 

By now, it should have become pretty easy to see how focus on getting the job finished economically, and on time, might get lost with the obvious potential for conflicts of interest.

A well-worn adage is that a project costs as long it runs, because so much of the cost is people, and people in this rather unreal environment are charged by the hour, at many many multiples of an average wage, even if doing straightforward jobs that might not attract much more than a labourer’s rate in the real world.  That doesn’t mean they get paid those rates of course.  But it explains why Crossrail wants another £450M and a further delay until 2022, though the trial running certainly won’t cost £450M, the hangers-on probably will.

Example Primavera project plan – the Waterfall at work

So-called Waterfall planning methodologies, as typically employed in mega projects, are no longer fit for purpose, and actually probably never were.  Plans are often created with a large number of copy-and-paste sections for lesser understood phases later in the project, before you know it, the project becomes huge, with matching costs.  Worse, each party involved creates one for their own “work package”, often in a silo.  A planner often has only passing knowledge of the actual work – they’re hired because they can drive Primavera “P6” – the de-facto planning tool, not for their detailed knowledge of software development for example.

Like the apocryphal algorithms of this summer, the output is inevitably only as good as the data that goes in, and there’s huge contractual and peer pressure to trot out highly detailed plans early on for the imagined production line, for things are are in reality a glimpse of the imagination – after all, the project is a prototype. These become the self-fulfilling prophecy, and contractors later congratulate themselves on “meeting the plan”. But on one nameless mega project, the planners (200 of them!) went offsite for a week to “replan the project”.  Not once, but every 3-4 months.   Because within 8-9 weeks of issuing the latest planning wisdom, it was embarrassingly far from reality.  If only they realised that they were in fact using an expensive sledgehammer to run a scrum/agile development methodology in little time-limited sprints that software developers are familiar with.

At its heart, a plan at the right level of detail does actually hold a possibility of success, because understanding the links – the dependencies – between activities, should, classically, become apparent.  The key to managing such projects is to minimise dependencies as much as is humanly possible, and then reduce more and more. That means going way above and beyond the normal planner remit to reduce dependencies between the silos, not as normally done trying to optimise one’s own silo.  The planning tools are designed to do exactly this, but a lack of high level hands-on planning resource, firm enough underlying plans, conflicting terminologies, and the herculean efforts required to tease out the real dependencies (which may not be at all apparent on the individual silo plans) means this is usually sketchy at best.  Worse, reducing the overall project dependencies may add what appears to be unnecessary cost, or cause a contractor to have to do things in a sub-optimal order for him.  That’s called “taking one for the team”, and of course somewhat unpopular, albeit good for the project as a whole.

2019 BEIS CO₂ comparison

Leaving aside short-term issues like the pandemic, pondering momentarily the possible long tail of effects on people’s revised working and travel habits, there’s little to beat saturation rail provision in cities at the moment, but rail might be more fairly seen as a seriously overfetishised 200 year old technology at significant risk of obsolescence.

The rail industry burnishes its image by the liberal application of greenwash, but it turns out that UK mainline railways aren’t very green at all.  BEIS figures show that 4 people in a modern car is as green as a typical UK train: sadly too many diesel-powered or at most 50% renewable powered, meaning the CO₂ figures shown are likely matched by other air pollution those aged diesel locos are certainly not Euro VI class! Further electrification has been scaled back, meaning the situation is likely to pertain for some time. There’s an interesting possibility of hydrogen powered electric trains to deal with the branch lines, but that’s looking a little way off as yet.  Contrast “domestic rail” with Eurostar, effectively nuclear powered via it’s French electricity and with few stops, numbers are very different.

This is also the possible green future for HS2 – we’ll ignore the forest destruction and the embodied carbon in all that concrete for now – no international comparator measures that yet either – but they may start to, and the bigger “total carbon” picture may turn out rather uglier.

The biggest risk to the rail industry are autonomous eco-friendly hydrogen-powered road vehicles taking you and your luggage from door to door in a reasonable and predictable time, which seems quite possible in the relatively near future – certainly when measured in glacial HS2 timescales – begging the question how much should now be invested in the 200 year old technology ?

Looking at these costs and timescales, compared to rapid advances in EV’s of various types and the looming Hydrogen Economy, it’s easy to see that if government were in a hurry to ‘clean up’ transport, new high speed railways could be one of the slowest and most expensive ways to do it.  Furthermore, EV and H₂ development is likely to accelerate much faster than rail for the above reasons, with the potential for rail to be left further behind.