Written evidence from Jonathan Tyler,
Passenger Transport Networks (HSR 138)
PREAMBLE AND
SUMMARY
1. Jonathan Tyler joined British Rail as a Traffic
Apprentice in 1962. His career path has been from operations through
demand modelling to BR-sponsored university lecturer and then
to independent-minded consultancy in a range of transport work.
Since 2000 he has specialised in making the case for the importance
of integrated strategic timetabling, drawing in particular on
Swiss methodology and software for case-studies in Britain. Most
recently he was invited by Greengauge 21 to design a joint timetable
for HS2 and the West Coast Main Line in order to illustrate the
opportunities for new services on the latter that would be afforded
by building HS2. He has also contributed to the service-planning
aspects of several other HS2 studies. It is this involvement in
the technical debate that has prompted the concerns expressed
in this submission.
2. The thrust of the submission is that the planning
process for new high-speed railways is too fragmented, based too
much on bold assertions and too little on wide-ranging analysis,
and too high-level and thus lacking in crucial detail to have
yet made a convincing case for building HS2 in the form presently
proposed. Attention is drawn to a troubling disconnect between
the frequency of services that is being promised and what can
credibly be delivered. However this is emphatically not an argument
against HS2 in principle, and certainly not an argument against
the construction of any new lengths of railway. Rather it is a
plea for a more rigorous, extensive and realistic approach, for
taking longer to examine the options but in the end reaching sounder
decisions, probably including a more even spread of benefits across
the regions, cities and towns of Britain. The paper restricts
itself to responding to selected questions in the Committee's
Inquiry; the author is aware of similar submissions that provide
more depth on some of the issues raised here. Full references
and additional detail appear in endnotes [p 8].
INTRODUCTION
3. According to the consultation document the
Government's vision is "a truly national high speed rail
network for the whole of Britain".1 That is a
desirable objective, but no specification of its structure, other
than of the "Y" sub-network, has been published. Instead
there seem to be unquantified assertions, generalised aspirations,
ill-defined commitments and neglected collateral effects. No obvious
mechanism exists for resolving the contradictions, and they provide
fodder for those who wish to block the proposals for quite other
reasons. In particular, no serious timetabling has been undertaken,
despite the fact that the timetable is the essential basis on
which a railway, of its nature, must be specified and operated.
DEFINING A
NATIONAL NETWORK
4. I take "national network" to mean
one that provides every urban centre with a reasonably equal quality
of service commensurate with its geography and size, supplemented
by suburban railways and bus connections. Only such a network—going
beyond a few big-city links—is capable of meeting future
demands for travel by a mode that is inherently sustainable in
a world where reducing carbon emissions and dependence on increasingly
expensive oil is imperative. Britain is some way from that ideal,
since service standards vary greatly, rail's modal share varies
greatly in consequence and governments, unlike those of some other
European countries, have never concerned themselves with defining
a programme to improve the situation.2 For the reasons
outlined in this paper the adopted scheme for High Speed Rail
does not provide a remedy either.
WHY WE
ARE WHERE
WE ARE
5. Before discussing the problems it is as well
to try to explain how we have got into this situation. The overriding
factors appear to be division of responsibility and differing
aims.
6. The Department for Transport [DfT] manages
the passenger franchises as largely separate units and with predominantly
financial objectives. Improvements to the infrastructure, however
welcome, are not explicitly influenced by seeking to meet national
standards of service and inter-urban connectivity. The commitment
to build new high-speed lines seems instead to be driven by broad
assumptions rather than by a detailed analysis of where the country
most needs capacity and accelerations. And local governments have
only a fragmented and reactive involvement, often, as in the case
of HS2, over-influenced by a determination to win a slice of the
action.
7. Responsibility for developing the proposals
has been devolved by DfT to High Speed Two Ltd [HS2 Ltd]. That
is not in itself an issue, but HS2 Ltd's approach is. It is clear
that the company is driven by the prospect of building the world's
most advanced high-speed railway, at the cutting edge of technology
and with dreams of exports. Whether that is sensible or appropriate
in a crowded island with dense development, relatively short distances
between urban centres and already-fast timings (at least on London
routes) has not been convincingly established.
8. Network Rail has the herculean task of managing
and improving the "classic" network and securing expansion
of its capacity. Nonetheless, in public at least, it appears semi-detached
from the planning of HS2, despite the substantial implications
of future inter-running between HS2 and classic lines. In the
same way Crossrail, another distinct entity, does not seem to
have had much involvement in HS2's plans for the interchange at
Old Oak Common.
9. Train operating companies mostly have short
franchises that render them uninterested in the long-term evolution
of services in their territory, and their responses to demand
are determined more by the imperative of profit than by any sense
of building a public network. They have contributed little to
the high-speed debate.
10. Finally, consultants engaged to develop schemes
are obliged to follow the specifications of their clients and
to maintain secrecy (out of concern for property values), with
the result that no national debate can be held until detailed
proposals are ready, by which time it is too late for the community
to have a proper conversation about the purposes and characteristics
of new railways.
THE FAULT
LINES IN
THE CURRENT
PROPOSALS
Confused strategies, poor forecasts and muddled
economics
11. Government documents repeatedly emphasise
the economic importance of faster inter-city railway services,
but it is curious how little analysis is presented of what the
real needs are if business travel is the primary motive. Instead,
without discussing the mix of business, commuting and leisure
generators of travel (the last has sub-categories of varying social
benefit), the argument tends to shift to the allegedly serious
impending shortfall in capacity.3 However this case
is flawed. It pays too little attention to weaknesses in pricing
tactics, to the economic implications of the disparity between
peak and off-peak volumes and to the scope for relieving the problems
(much sooner than HS2 could do) by lengthening and redesigning
trains, by comparatively modest infrastructure works and by a
thorough review of over-cautious operating practices.4
12. Three particular features are of concern.
One is that classical economic arguments in favour of widening
employment markets (to the point, for example, of encouraging
mass commuting between Milton Keynes and London and the growth
of Birmingham <> London commuting) will increasingly be
tempered by resource and carbon constraints. Such travel may even
be greatly curtailed. The second is that a significant element
of the statistical evidence of overcrowding comes from operating
companies who have an interest in talking up the problem and the
accuracy of whose data is not publicly tested. And third, the
Government's normal economic rigour should be applied to questioning
whether the very high cost of alleviating the peak of the peak
out of Euston on Friday evenings can be justified.
Maximum speed
13. HS2 Ltd has specified a railway with a maximum
speed of 400 km/h. That is beyond the range of 300 to 360 km/h
that is presently typical of new lines and trains. Given the disproportionate
increase in energy consumption as speed increases and the effect
of very high speed on route geometry one would have expected more
open and deliberate analysis of the costs and benefits of options
for, say, 300, 350 and 400 km/h. For example, between Birmingham
and London running at 300 km/h would add only about five minutes
to the 400 km/h schedule of 49 minutes: it is not intuitively
obvious that that would make much difference to rail's ability
to capture traffic.5
Station location
14. The consultation paper emphasises the role
of the railway in serving city centres.6 This has been
a fundamental tenet of the design of railways throughout their
history, and it has become normal European practice where high-speed
services have been introduced. Moreover, likely trends make central
stations critical: the new thinking on place-making clearly indicates
this, while building ex-urban stations that principally depend
on cars for access would be socially inequitable and (at best)
unwise in terms of sustainability.
15. Yet the present proposals refer imprecisely
to what appear to be "parkway" stations for South Yorkshire
and the East Midlands. Equally, the service specifications suggest
that the preoccupation with fast services between major centres
has led to the omission of calls at places such as Carlisle (a
railhead serving a huge area) and Stoke-on-Trent (desperately
in need of regeneration).
Connections with the "classic" railway
16. Closely associated with the economic-geography,
access, land-use and resource factors in favour of central stations
is the question of connections with regional and local public
transport networks. These focus on main stations since their own
primary function is to serve the concentrated activity of city
centres and since that optimises their links with longer-distance
services. If the links are broken because high-speed trains serve
separate stations then swathes of the population will be denied
a share in the benefits from what is likely to be a Government-funded
project. A two-tier system of access to fast transits would be
economically detrimental.
17. HS2 Ltd has adopted a seemingly cavalier
approach to the impacts of its proposals on some users (notably
those from Coventry)7 and on operation of the classic
network. For example, just north of the junction between HS2 and
the West Coast Main Line [WCML] near Lichfield lies one of the
latter's critical constraints, namely the flat junction at Colwich
and a two-track section in a four-track railway. It is proposed
to add a significant number of extra services, at least during
phase one: it is not that they cannot be accommodated, rather
that to do so will have a material, yet barely acknowledged,8
effect on the entire WCML (and hence the national) timetable.
18. Similarly, HS2 Ltd holds out the prospect
of connections between its trains and Heathrow Express and long-distance
services on the Great Western Main Line [GWML] at a multi-function
interchange at Old Oak Common in west London. The Heathrow case
derives from the assumed need to include the airport in the high-speed
scheme and is more persuasive than the idea of a direct spur [on
which see paragraphs 24-25]. The longer-distance case is arguable,
since the territory that could benefit is quite limited. GWML
is already operating close to its limits, and even when the pending
electrification and resignalling add capacity that will quickly
be absorbed by indigenous growth. Apparently unaware of this,
HS2 does not appreciate that stopping just the Heathrow trains
would remove the performance buffer and outer-suburban paths,
while the only practicable timetable including stops in GW services
would require every train to call (because that, and twin
platforms, is the only way to maintain a smooth flow). This would
lead to costly works and cause an unacceptable loss of time for
GW travellers, relative to marginal gains to interchangers.
Incoherent timetabling of HS2 services
19. Incoherence in timetabling is also evident
in questions about how HS2 itself would work. No evidence has
been published about the effect on headways of intermediate stops
at Birmingham Interchange and Old Oak Common, apart from unspecific
statements about track geometry. Operation of a very frequent
service imposes particular requirements on the design of the new
Euston, yet turnround and platform-reoccupation times seem optimistic,
and it is doubtful whether space exists for the grade-separation
that may be essential to avoid conflicts at the "throat".
And published summaries of the proposed services are disjointed.
We discuss the links to HS1 and Heathrow and the matter of the
total number of paths in separate sections below; here we draw
attention to issues concerning the pattern of services and the
treatment of particular places.
20. Carlisle and Stoke-on-Trent were mentioned
above. The descriptions of the modelling hardly refer to how sub-regional
centres will be linked to the core network. No strategy has been
set for through running or systematically-planned connections
to serve places such as Wakefield, Huddersfield, Rotherham, Chesterfield,
Derby, Motherwell, Bolton and Wolverhampton, and no data has been
proffered on the share of total demand and potential economic
benefits that such places represent or on whether HS2 trains could
be adequately loaded without their traffic.
21. Diagrams outline possible services between
provincial centres on the two arms of the "Y" route,9
but those on the eastern arm are meaningless so long as the location
of the stations is unknown (greenfield sites would have negligible
value for intra-regional journeys) and those on the western arm
raise difficulties in serving Macclesfield, Stoke and Wolverhampton
which in turn affect the working of WCML in ways that are being
ignored. And it is very odd timetabling to suggest that all four
Birmingham trains, none of the Manchester trains and one random
Leeds train will call at Birmingham Interchange.
LINKS TO
HS1 AND LONDON
HEATHROW AIRPORT
22. The Government believes that HS2 should be
connected with HS1 (the line from London St Pancras to the Channel
Tunnel and the continental railway network) in order that through
trains can operate between British and key European cities. The
cost of such a link (including border controls at regional stations)
would be considerable, its operation would be extremely difficult
(it has not been shown that it is even feasible to timetable HS
trains along a section of high-frequency urban line), and it would
import into the working of HS2 just the kind of perturbation that
HS2 Ltd, when specifying frequencies, has said must be avoided.
In any case, the pattern of services could not possibly offer
particularly attractive frequencies in competition with direct
air services.
23. The initial evaluation of demand for through
services demonstrated that they could not be justified and proposed
a local connection between Euston and St Pancras (which would
not entail any greater inconvenience than travellers endure at
airports). The Government has chosen instead merely to assert
the need for the link,10 abetted by parts of the railway
promotion lobby that fantasise about romantic international expresses
but hardly analyse the market.
24. A similar situation exists in respect of
the proposed link to Heathrow. Several evaluations have found
against it on the grounds of demand relative to the cost and to
the practical difficulties of securing an effective service without
degrading the London offer—the prime purpose of HS2. However
the idea of a link is powerful if one buys into belief in an ever-expanding
global economy without noting the probability that the very environmental
constraints that are used to justify transfers to rail are likely
to curtail aviation, especially business trips for which video
conferencing is a viable substitute and the marginal leisure trips
which cannot conceivably provide an economic rationale for public
expenditure on so expensive a new piece of railway.
25. The plan for an interchange with Crossrail
(but not Heathrow Express [see paragraph 18]) is a plausible alternative
with other strong arguments in its favour (though it does depend
on acceptance of the westerly route though the Chilterns), but
this is seen only as an interim measure (at great cost), partly
on the basis of the government's visceral objection to interchanging.11
That is not a convincing foundation for serious transport planning.
THE CAPACITY
OF HS2: ASPIRATIONS,
PATHS AND
THE URGENT
NEED FOR
REALISM
26. Given the Government's presentation of HS2
as a national project and given the strong sense in the regions
(especially in Scotland and northern England) that they should
obtain early and tangible benefits from this huge investment it
is entirely understandable that many aspirations are being expressed
for inclusion in the pattern of high-speed services. At the same
time HS2 Ltd has uncritically bought into the story about capacity
promulgated by Network Rail and Virgin Trains and thus extrapolated
current WCML services to 2026 and HS2 without sufficient analysis
of priorities. The result is an extraordinary muddle that threatens
the credibility of the entire scheme. The issues are summarised
in the table overleaf (note that the numbers of paths are for
the peak period, since that is what matters most when planning
the timetable).
27. A railway has a finite capacity that depends
on the layout and speed of the route, the design of signalling,
its operational rules, the acceleration, braking capability and
stopping patterns of the trains and the mix of these characteristics.
As a general rule, the more homogeneous the services the more
trains can be operated. High-speed lines tend to have this property,
but conversely headways must be wide enough to secure safe separation
between trains in the event of a serious incident. At present
no high-speed line anywhere in the world runs at more than 14
trains/hour.12 Advances in signalling may raise this
to 16 as a peak-period maximum, with some performance risks accepted
and if fewer paths are used off-peak to provide a recovery margin.
28. HS2 Ltd believes that 18 trains/hour may
eventually be feasible, but many railway people see this as supremely
optimistic, dependent on far-from-proven technology and thus an
unsound base for planning: it would be irresponsible to raise
expectations and rest calculations of benefit on unachievable
scenarios. Moreover, it is expressly admitted that, while 18/hour
allows for everyday, unavoidable perturbations within the high-speed
system, it would by that same token only be practicable in circumstances
where the system has been largely quarantined from the more troubled
conditions on the classic railway. The fact that that may not
be realised for many years and may be undesirable, for reasons
argued earlier, thus strengthens the case for caution.
29. However HS2 Ltd now envisages 18 for the
Y network without explaining why it thinks that will be achievable
by the opening of phase 2.13 The proposed distribution
of paths is shown in the first column of the table. Broadly the
plan is to run four Birmingham services, six on the eastern arm
of the Y and eight on the western arm. A strange note is then
appended: "Further work is being done to determine which
of the above services might serve Heathrow and which might run
on to mainland Europe". That is more than a technical matter
of "further work"! It is absolutely fundamental to the
concept and objectives of HS2.
30. Column 2 sets out a specification that combines
the aspirations expressed around the country with an outline evaluation
of the ideal frequencies required to make the services attractive
in terms of convenience (assuming continued availability for the
majority of travellers of the ever-flexible car). The total is
an impossible 27 to 30, including provision for HS1 and Heathrow
trains. Something must give, and an optimistic maximum scheme
is presented in column 3. This cuts Edinburgh and Glasgow to alternate-hour
trains each, removes a Newcastle service, sharply trims the Yorkshire,
East Midlands and North West patterns and reduces both HS1 and
Heathrow to a thin provision of only two trains/hour each. Column
4 removes two more paths to bring the total to an altogether more
rational planning maximum of 16 trains/hour.
31. At that point it must be asked whether the
listed HS1 and Heathrow services would offer sufficient flexibility
to attract travellers who have the alternative options of either
direct flights or a relatively simple interchange along the Euston
Road. Operationally too the need to provide connections would
impinge on timetable planning. And above all the frequency could
not justify the cost of constructing the two links while the unavoidable
loss of Euston services would undermine the economics of HS2,
with consequences probably spreading back to the capacity that
it was supposed to release on WCML. The inescapable conclusion
is that links to HS1 and Heathrow cannot be justified and should
be removed from the Government's proposals. They are simply not
credible.
32. The outcome of achieving clarity on this
issue is shown in column 5, where the freed paths are redistributed
to give a more appropriate domestic network: these are the services
timed in detail in the Greengauge exercise in proposals wholly
integrated with other services on WCML.14
33. There is one possible source of relief, but
it is another sign of the confusion surrounding pathing that statements
differ and practical proposals are absent. HS2 Ltd proposes that
the Birmingham <> London trains should be composed of two
200-metre sets, and the stations are to be designed to accommodate
400-metre trains. Yet the trains working through from and to the
classic railway would be formed of single 200-metre sets (presumably
total separation might allow coupled units for Leeds and Manchester
eventually, but not for other places). This would represent substantial
under-utilisation of capacity.
34. The solution of course is to run separate
sets between provincial centres and locations in the Midlands
where they would couple to run over the core HS2 section - and
divide on the return.15 The practice is commonplace
with German and French high-speed trains, which seem to manage
it effectively. The suspicion must be that HS2 Ltd is so committed
to achieving the fastest possible times between major centres
that it has lost sight of practical service-planning. Joining
and splitting does require time for the technical operation, a
performance margin and one or two stations on the existing railway
rebuilt to handle 400-metre trains, but it would bring with it
great gains in connectivity and frequency and more effective use
of the new capacity.
35. The implications are shown in the last two
columns of the table. It could be used to retain two trains/hour
for both HS1 and Heathrow, but it would be better (column 7) to
maintain a worthwhile frequency on each Y arm and to run only
14 trains/hour. Natural locations for joining and splitting are
Derby and Stafford. Crucially, pursuing this concept implies closer
integration between the old and the new systems than HS2 Ltd might
wish—but with large potential benefits—and it would
predicate comprehensive timetabling of the whole network.
36. Finally a point must be made about regulation
and competition. The analysis of possible distributions of paths
assumes a planned railway (because a complex system demonstrably
depends for its optimal functioning on planning). If multiple
operators are granted access by a regulator to compete for the
most lucrative markets it is certain that services for the less
lucrative markets will be cut back, to the detriment of an even
spreading of benefits. The ambiguity in government policy on this
matter must soon be resolved. It should be noted that a service
planned by a public-interest agency does not exclude its delivery
by a range of contracted parties.
WHAT SHOULD
NOW BE
DONE?
37. The strategic objectives of HS2 are in disarray.
Timetable planning on the existing network is not in good shape,
as the East Coast saga has shown.16 Awkward interfaces
between plans across the country are becoming more evident, and
none so much as those between HS2 and the classic railway. It
has not been proven conclusively that the capacity shortfall on
WCML merits a huge single project that cannot deliver benefits
for fifteen years, in preference to enlarging the programme of
smaller, more immediate incremental projects across the country,
particularly ones that might appreciably lift rail's very low
modal share. And timetabling is emphatically not some secondary
technical activity that can be undertaken long after large projects
have been designed.
38. In my view, what is needed is a process that
(a) starts with a matrix of flows by all modes and models future
demand under a range of scenarios; (b) sets targets for modal
shares driven by environmental (eg. carbon-reduction) as well
as economic factors; (c) derives a national timetable plan to
meet the forecast demand, within a framework of standards and
free of historical and institutional preconceptions; and (d) draws
up a national rail infrastructure plan to enable the planned services
to run in an operationally efficient manner. Such a process might
well generate a case for new lines, including even something akin
to the current proposal, but they would be embedded in a coherent
plan for the benefit of the country as a whole. The tools and
the data exist for such an exercise—and to undertake it
would bring Britain into line with Switzerland, whose much admired
public transport network has been built through applying broadly
this methodology over the last 30 years and which already enjoys
a vision for the next 20.
May 2011
1 Department for
Transport [DfT] (2011 - February). High Speed Rail: Investing
in Britain's Future. Consultation. This quote [at p.28] comes
from the May 2010 Programme for Government.
2 For one of countless
examples see: Abbott, S (2011). A lament for Lincoln. Modern
Railways, May, pp 69-73.
3 For example,
the consultation document [op cit, pp 9-10] slips from
generalisations about economic growth and competitiveness into
an uncritical discussion about capacity, for which high-speed
lines may, or may not, be the solution. The too-easily-forgotten
Eddington Transport Study [Department for Transport (2006)] concluded
[Volume 2, para 2.18] that "average rail journey times between
major UK urban areas are close to that [sic] of other European
countries, with journeys between London and other UK major cities
performing particularly well relative to journeys from other European
capitals". In its advice to Government [Executive Summary,
paras 1.50-1.52] it placed the emphasis on the density of transport
demand rather than on high speed. The genesis of the present proposals
lies in the former but has been taken over by the latter. The
report is archived at
http://webarchive.nationalarchives.gov.uk/+/http://www.dft.gov.uk/adobepdf/187604/206711/executivesummary.pdf
and ~ /206711/volume2.pdf.
4 Some safety
rules are unnecessarily restrictive, some timing margins are longer
than they need be in order to protect Network Rail's performance
record, and on some sections more trains could probably be run
in peak periods by trading off the advantage of more seats against
occasional risks to punctuality.
5 Author's estimate
from the published diagrams
[www.dft.gov.uk/pgr/rail/pi/highspeedrail/proposedroute/maps/].
Comparisons between current and high-speed journey-times must
be transparent with regard to recovery margins: the latter are
probably being quoted with modest allowances whereas the former
are inflated by the malign effects of the prevailing performance
regime.
6 "Inter-city
rail links have an unrivalled capacity to enable rapid and direct
journeys between central business districts" [op cit,
para 1.32]; HS2 Ltd's "demand-led" approach indicated
"city centre station locations with high quality onward transport
links" [para 4.6].
7 It described
its future London service of one train/hour (rather than three)
as "residual" [DfT (2009). High Speed Rail: London to
the West Midlands and Beyond. HS2 Technical Appendix, Appendix
2, p 17].
8 "It is
assumed that some infrastructure/signalling works have taken place
in the Stafford area to alleviate this known capacity constraint"
[ibid, para 2.20]. Any project would be complex, environmentally
fraught and very costly.
9 DfT (2011—February).
Economic Case for HS2: The Y Network and London—West Midlands.
Figure A2, p 61.
10 Compare the
two appraisals: DfT (2010). High Speed Rail: London to the West
Midlands and Beyond. A Report to Government by High Speed Two
Limited*, section 3.8, and paras 3.24-3.33 of the Consultation
document [op cit].
* http://webarchive.nationalarchives.gov.uk/20110131042819/http://www.dft.gov.uk/pgr/rail/pi/highspeedrail/hs2ltd/hs2report/pdf/chapter3d.pdf
11 The Secretary
of State is reported to have said "Lug your heavy bags down
a couple of escalators along 600m of corridor and then change
trains at a wet suburban station somewhere in north west London.
That is not an option"
[http://www.estatesgazette.com/blogs/jackie-sadek/2010/08/transport-sec-chops-down-old-oak-high-speed-2.html].
12 As far as this
author is aware. The Tokaido Line runs at 14 trains/hour in the
morning peak into Tokyo. There are plans for 15/hour in France
after 2020.
13 See the figure
referenced at endnote 9. The original specification was 14 trains/hour
in the peak (10 off-peak) with "an ultimate capacity of 18"
[see endnote 7, Appendix 1, para 2.3.2].
14 Greengauge21
(2011). High-speed rail: capturing the benefits of HS2 on existing
lines. The present author designed the timetable summarised in
this report; a detailed technical report will shortly be available.
See:
www.greengauge21.net/wp-content/uploads/Capturing-the-benefits-update.pdf.
15 The consultation
document refers rather casually to portion working for Heathrow
services [op cit, p 66], and two associated Factsheets
mention coupled running, but no other reference is known. See:
http://highspeedrail.dft.gov.uk/sites/highspeedrail.dft.gov.uk/files/hs2-trains_0.pdf
and ~ files/connecting-to-heathrow_0.pdf.
16 See: Tyler, J (2010).
How not to write a timetable. Modern Railways, November,
pp 64-66.
| 
