Memorandum by the Passenger Transport
Executive Group (OPT 14)
OVERCROWDING ON PUBLIC TRANSPORT
INTRODUCTION
1. The Passenger Transport Executive Group
(PTEG) has substantial evidence of overcrowding on rail services
within certain PTE areas causing significant hardship to passengers.
PTEG believes that overcrowding on the heavy rail network is a
result of a failure to plan for growth due to inadequacies in
the franchise mechanism, poor operational performance, failures
in the rolling stock leasing market and the lack of a clear rolling
stock strategy.
2. The evidence below relates to recent
experience of overcrowding on PTE rail services in England. It
particularly relates to experiences in West Yorkshire, the West
Midlands and Greater Manchester. Evidence is also provided relating
to problems of overcrowding on the Manchester Metrolink light
rail system.
BACKGROUND
3. PTEG represents the six English PTEs
and Strathclyde PTE in Scotland. The PTEs cover areas with a population
of 13 million people. The GDP in the six English PTEs constitutes
approximately 20% of the GDP for England. PTEs are responsible
for the investment of large amounts of public money (during 2001-02
PTA/Es collectively were responsible for almost £1 billion
of investment in transport) and the majority of this is in the
form of indirect and direct subsidy support to the commercial
bus and train operators. The PTE areas include the busiest stations
and the most heavily used local rail services outside London.
All PTEs are co-signatories to the appropriate local rail franchises
in their area and, therefore, feel well placed to comment on the
impact that overcrowding is having on the ability for rail to
play a full role in an integrated transport strategy.
4. PTEs are committed to delivering high
quality integrated transport networks for the conurbations they
represent. Through investment in rolling stock, new stations and
other improvements to the network, use of the local railways in
conurbations has grown significantly. Furthermore, each PTE has
in place Local Transport Plans and Rail Strategies for further
developing their networks. These strategies recognise that if
rail is to continue to play its part in an integrated transport
network for the conurbations, the service offered must be of a
significantly consistent high quality.
5. Market research indicates that one of
the most important issues for passengers after reliability, punctuality
and security/safety is overcrowding whether on longer distance
or metropolitan services. Passengers in different areas do, however,
have different tolerance levels depending on the nature of the
rail services and the alternatives available. If PTEs are to make
a contribution to national policy objectives (including the 50%
growth target in the 10-year plan) and also meet their own Local
Transport Plan objectives they need to ensure that rail is a real
competitor with the private car in terms of frequency, journey
times, comfort and ambience.
EVIDENCE OF
OVERCROWDING
6. There is clear evidence of significant
overcrowding on several PTE rail networks. In West Yorkshire,
passengers have had to endure persistent overcrowding at peak
times. On no single day since the start of the franchise in 1997
has all the planned peak capacity been provided. Since the start
of the franchise the average achievement rate of planned peak
train formations has been less than 80%. Put another way, a regular
commuter can expect, on average, to experience a train formed
of less than the booked number of carriages twice a week. Overcrowding
is often of such a scale that passengers either cannot board trains
or have to endure conditions so cramped that the journey is extremely
unpleasant.
Example 1
The 06.52 York to Leeds via Harrogate train
is a key commuter service into Leeds arriving at 08.20. The average
number of passengers using this service in autumn 2002 was 245.
The train is planned as a three-car pacer unit with 195 seats
meaning that the average load factor is 126%. Pacer units are
unsuitable for commuter services as they have narrow bus-style
seats and aisles reducing the availability of standing space.
Between the start of the winter 2002-03 timetable
and the middle of December 2002, the train has only had its correct
three-car formation on 43% of occasions. That is to say on more
than half the days in question the train operated as a two-car
train. In the majority of cases, the substitute unit was a two-car
sprinter with approximately 150 seats, which implies a load factor
of 163%. On seven occasions the train operated as a two-car pacer,
which implies a load factor of more than 200%, but in reality
some West Yorkshire passengers would be unable to board the service.
7. Within the Centro area in the West Midlands,
overcrowding is perceived to be a significant issue among many
peak rail users, and it is not unusual for passengers to complain
about not being able to board trains due to the severe overcrowding.
One in six commuter journeys into Birmingham is made by rail,
but increasing overcrowding will create a significant barrier
to growing this proportion. Market research shows that most rail
users have a car available for their journey, but choose to use
the train. Unless rail users can be offered a comfortable service,
including having a seat, then many will choose to drive, further
congesting the road network. Under these circumstances, encouraging
existing car users to transfer to rail becomes virtually impossible.
8. Within Greater Manchester, regular overcrowding
on the heavy rail network occurs on the Bolton-Manchester, Greenfield-Manchester
and Atherton lines at peak times. There are also overcrowding
problems on the Metrolink light rail system and these are evidenced
separately towards the end of this submission.
CAUSES OF
OVERCROWDING
9. Overcrowding occurs on both a planned
and unplanned basis. Planned overcrowding is a result of an inability
to match the available rolling stock resources with passenger
demand. Unplanned overcrowding results from a deviation from the
planned train service and train formations on a day-to day basis.
The main causes of overcrowding problems in PTE areas are:
— Planned capacity is insufficient
to meet demand;
— Shortfalls in delivery of agreed
capacity caused by:
(a) a shortage of rolling stock;
(b) inadequate penalties for under-delivery;
(d) A lack of resilience.
CAPACITY PLANNING
10. Fundamentally the original franchises
did not make adequate provision for passenger growth and the consequential
need to provide additional capacity. In short, growth of the levels
achieved in some PTE areas was either not envisaged or simply
not properly planned for in the original franchises. Where growth
has occurred, the franchise mechanisms have been unable to fully
respond to it. For example, in West Yorkshire, peak patronage
increased by over 40% in the first three years of the franchise
and there was no corresponding increase in rolling stock capacity.
11. Capacity and overcrowding are measured
through a capacity regime. Different regimes operate in PTE areas
compared to those operated in the South East. There are several
reasons why different overcrowding standards need to be applied
in PTE areas:
— rail demand outside London and the
South East is still relatively elastic. The conditions on the
highway networks, although congested, are not as severe as in
London. If rail is to effectively compete with road car drivers
need a reasonable expectation of a seat and personal space. If
their experience of rail is unfavourable, they will switch back
to car;
— much of the rolling stock deployed
in PTE areas is unsuitable for carrying high volume peak commuter
flows. In some areas, a large number of services are still provided
by pacer and sprinter units. Pacers with their few external doors
and narrow seats and aisles cannot physically accommodate large
volumes of standing passengers. Most sprinters were designed for
longer distance flows with 2+2 seats and doors at each end of
the vehicles again not suitable for high volume commuter flows
with large numbers of passengers attempting to board in a short
space of time;
— many PTE services are operated by
an intricate mix of rolling stock types (for example in West Yorkshire
there are 10 different unit types all with varying capacities).
Train operators and PTEs have developed train plans that optimise
the use of scarce rolling stock resources by matching train loadings
as closely as possible to unit types which are often deployed
in combination. The result is often a train plan that is not robust
enough to consistently deliver the planned capacity as perturbations
(for example due to unit breakdowns) cause the substitution of
a lower capacity unit and the inevitable overcrowding. In the
West Midlands, electric services are limited to either three or
six car formations which means that where a four car train is
required either an overcrowded three car, or underused six car
train has to be planned;
— Peak loadings can vary significantly
from day-to-day and therefore on any given day actual loadings
and overcrowding can be significantly higher than the train plan
suggests.
— PTEs need to be able to offer a consistently
high quality rail service in order to meet Local Transport Plan
objectives and to support regional economies.
12. In West Yorkshire, planned capacity
is measured through the Passengers In Excess of Capacity (PIXC)
measure defined in the Franchise Agreement. PIXC is an aggregate
measure of the proportion of passengers travelling in the peak
period in excess of the number of seats provided on each train.
The Franchise Agreement states that PIXC for West Yorkshire should
be no more than 4.0% in the morning and evening peak periods.
13. At present PIXC in West Yorkshire is
9.0% in the morning peak (more than double the permitted level
in the Franchise Agreement) and 5.3% in the evening peak (lower
because it covers a longer time period). This means that there
are a significant number of overcrowded trains even when the service
operates correctly.
14. Within the Central Trains franchise,
Central Trains is required to deploy up to 112 vehicles on PTE
sponsored rail services. The franchise as written requires that:
(a) Passengers shall not be required to stand
on trains for more than 15 minutes during peak periods and not
at all at any other time.
(b) To the extent that passengers are obliged
to stand during peak periods the number standing shall exceed
the seating capacity of the train by no more than an average of
10% on Weekdays, provided that the number standing shall exceed
the seating capacity of any individual train by no more than 35%
on any individual service on any day.
15. In the West Midlands, overcrowding is
monitored on an individual train-by-train basis rather than the
aggregate PIXC approach used elsewhere. Passenger counts are regularly
undertaken on all peak trains. If a train has, on average, a load
of more than 110% of its seating capacity, then the train operator
is required to provide additional capacity.
16. The obligation on the train operator
to meet the overcrowding standards in the Franchise Agreement
only extends up to a specified number of vehicles. In the West
Yorkshire this cap was actually reached from day one of the franchise.
Therefore there has been no obligation whatsoever on the franchise
operator to fund additional vehicles during the course of the
franchise. This inadequacy in the franchise has been the fundamental
cause of overcrowding problems experienced in West Yorkshire.
17. In the West Midlands, the rolling stock
cap was reached a couple of years ago, but overcrowding has been
kept within the 110% limit until this year (in part due to other
longer-distance operators being able to carry significant volumes
of commuters). The current position is that unless a way of funding
additional rolling stock is found, passengers will have to suffer
conditions worse than envisaged in the franchise agreement.
18. PTEG believes that the SRA should fund
rolling stock to meet capacity standards through the franchise
mechanism. Due to the failure of this process some PTEs have developed
Rail Passenger Partnership (RPP) bids on a piecemeal basis to
try and bridge the gap. In Greater Manchester, funding was made
available to retain some pacer units that were due to go "off
lease", but the impact has been negated by problems in introducing
new rolling stock.
19. Metro has sought and received RPP funding
totalling £6.3 million for additional rolling stock in West
Yorkshire. Bids have been prepared at considerable cost to Metro,
but it was the only mechanism available to fund additional capacity
that was desperately needed. To date an additional 13 diesel vehicles
and 16 new electric vehicles have been funded for use in West
Yorkshire through this mechanism. The impact of the additional
diesel vehicles has been reduced by an underlying shortfall of
rolling stock in the franchise. The additional electric vehicles
have allowed half of the fleet of new Class 333 units to be strengthened
to four-car units. The remainder of the fleet will be converted
in 2003.
20. The result of additional rolling stock
provided through the RPP mechanism has assisted in reducing the
growth in overcrowding, but not solved the problem per se. As
noted above, PIXC in West Yorkshire is still twice the permitted
limit in the morning peak and this does not take any account of
the day-to-day shortfalls in the provision of planned capacity.
SHORTFALLS IN
THE DELIVERY
OF PLANNED
CAPACITY
21. Shortfalls in planned capacity are compounded
by failures to deliver on the day where train operators are unable
to provide the planned capacity.
Shortage of rolling stock
22. The main cause of unplanned overcrowding
in West Yorkshire is a fundamental shortage of rolling stock to
operate the services. In February 2001, the Arriva Trains Northern
(ATN) franchise was extended for a period of two years and as
part of the contract, an underlying shortfall of 18 vehicles (9
x two-car units) was identified. The shortfall manifests itself
most often in short-formed peak trains.
23. To date, four additional vehicles (two
x two-car units) have been sourced by ATN (and funded by the SRA),
but this still leaves an underlying shortfall of 14 vehicles for
which there is no readily identifiable source. Even if they could
be sourced, the SRA has indicated that its current budgetary position
means that it would be unable to fund them.
24. There is currently a national shortage
of rolling stock with no readily available rolling stock that
can be drafted in to meet shortfalls in specific areas. The only
real solution is new build, but the long lead times and incompatibility
with existing rolling stock often rules this out as a short term
option.
25. The timescales associated with procuring
additional rolling stock are also an issue as suitable spare rolling
stock does not currently exist. The lack of a national rolling
stock strategy means that re-allocation of rolling stock between
franchises to meet particular shortfalls is difficult (except
between franchises with the same parent operating group).
Economics of rolling stock
26. The difficulty with procuring additional
rolling stock is the costs involved which create a significant
ongoing need for subsidy. The basic economics are that it costs
about £150,000 per year to lease, maintain and operate each
vehicle. Given that there is plenty of spare capacity off-peak,
this cost effectively has to be justified on the basis of one
morning and evening peak journey. The maximum income that a single
peak round trip is likely to generate would be around £50,000
assuming a full and standing load. Therefore every additional
peak vehicle needed to reduce overcrowding would generate a subsidy
requirement of £100,000 per annum. This is a significant
difference from the South East where longer journeys mean that
passengers pay more per journey and thus the direct lease and
operational costs can often be covered. Under this scenario PTEs
are very concerned that London and the South East will be seen
as better value for money for investment in rolling stock, and
reducing overcrowding in PTE areas could be considered unaffordable
by the SRA.
27. The high cost and unavailability of
rolling stock is a clear indication of market failure that the
Rolling Stock Leasing Companies have failed to address. The few
"speculative" orders placed for rolling stock have in
fact quickly filled existing gaps. There is also no short-term
leasing market for the type of rolling stock that could be readily
deployed on commuter services.
28. Rolling stock leasing companies are
generally seeking lease lengths beyond the end of franchises in
order to reduce uncertainty (and consequential risk pricing).
This is particularly problematic for franchises that are close
to their renewal dates (or in the case of Arriva Trains Northern
about to enter an "interim" two-year franchise period).
The rolling stock leasing companies are risk averse and the SRA
is reluctant to provide the necessary ongoing commitments indicating
a failure of the rolling stock market. Shorter franchise lengths
could exacerbate this problem.
Incentive Regime
29. Where PTEs have a capacity regime in
place, shortfalls in agreed capacity are measured through the
Short Formation Incentive Payment (SFIP) regime. Each peak train
contained within the train plan has a capacity based on the number
of seats to be provided. Any deviation from this is recorded as
an SFIP failure and a penalty is levied based on the difference
between the actual and planned number of seats provided.
30. SFIP penalty payments are related to
the main Performance Incentive Regime (and must continue to be
so in order to prevent perverse incentives to cancel a train rather
than short-form it). However, the penalties do not currently relate
avoidable cost of leasing an additional unit of rolling stock.
The West Yorkshire example below indicates that there could be
an incentive to short-form rather than lease additional rolling
stock (where there is an underlying short fall such as in West
Yorkshire).
Example 2
The 07.15 Sheffield to Leeds service forms a
key commuter service into Leeds arriving at 08.30 carrying an
average of 290 passengers. The train is booked to operate as two
pacer units coupled together which would provide 243 seats implying
that 47 passengers would have to stand. Since the start of the
winter 2002-03 timetable the train has only operated with the
booked capacity on 55% of occasions. The other 45% of occasions
it has generally operated as a single two-car unit implying a
load factor of up to 240%.
The SFIP penalty for such a failure is in the
order of £200 per day or £50,000 p.a. This can be compared
with the annual lease cost of an additional two-car unit of at
least £200,000 p.a. (excluding maintenance and staff cost).
31. In West Yorkshire, the SFIP regime does
not apply to some of the busiest commuter routes (including Harrogate
to Leeds and TransPennine services between Huddersfield and Leeds)
and therefore no penalties are payable for short-formations. Some
of these routes have some of the worst records of short formations
and overcrowding. On the TransPennine route for example one morning
service between Huddersfield and Leeds loads to 184% of seating
capacity and is regularly formed of less than the booked number
of carriages.
Service Disruptions
32. Service unreliability is also a significant
cause of overcrowding. Recent exceptionally poor performance in
the West Midlands has meant that overcrowding is a frequent occurrence
across the network and is the single largest cause of overcrowding.
Overcrowding occurs when, for example, the cancellation of a train
results in a double load on the next one, or late running means
that some of the load of the next train is picked up.
33. Overcrowding can be both the cause and
the effect of late running. A shortformed train that is seriously
overcrowded will spend longer in stations while passengers board
and alight, and thus it will run late. On a high frequency local
service, any late running will mean that it will pick up passengers
arriving for the next train, resulting in further overcrowding
and late running.
34. Problems such as this are a particular
problem in the morning for passengers trying to catch trains from
stations closer to the city centre, as by the time the train reaches
them they are often unable to board.
Lack of Resilience
35. The original franchises were let on
the basis of a reducing cost base. This led to operators making
efficiency savings or alternatively working existing resources
more intensively. In many cases assets have been worked too hard
and there is no resilience left in the system.
36. The current peak train plans rely on
very tight resource diagrams in order to maximise the usage of
rolling stock. However, this means that the timetable is very
tight and prone to delay. Building more slack and turn-round allowance
into unit diagrams would have performance benefits, but would
result in greater overcrowding as less efficient use of the rolling
stock is being made. There is therefore a trade-off between performance,
overcrowding levels and costs.
37. In West Yorkshire, there are also specific
examples of staffing levels being reduced, which has had a direct
impact on overcrowding. For example, staffing levels were reduced
in the Service Delivery Centre (by the original franchisee, MTL)
such that controllers were unable to sufficiently focus on allocating
the correct unit to each diagram, particularly during `out of
course' running when the focus is necessarily on keeping the service
going. Staffing has since been increased again by Arriva Trains
Northern, but this serves to illustrate the false economies of
staff reductions.
OTHER ISSUES
38. Another issue which impacts on overcrowding
is the fact that new or refurbished trains now have to comply
with the Rail Vehicle Accessibility Regulations. In order to make
the trains more accessible for disabled people, seating capacity
is inevitably reduced. Thus if an existing fleet of trains is
refurbished, there will be less capacity than before and either
more vehicles would be needed, or greater standing would have
to be tolerated.
39. Many PTE routes are shared with other
operators. By planning jointly with other operators, good use
of overall rolling stock can be made. However, if another operator
changes the times of their trains, then this can significantly
alter the loading of the PTE services in a way that may not be
manageable. Depending on other operators to carry local passengers
also means that it is not possible to impose the PTE loading standards
on these services.
40. For example, one of the most overcrowded
services in the West Midlands is Chiltern Railways 1710 Birmingham
Snow Hill to Marylebone service which has an approximate 170%
load factor, however Chiltern consider that tackling overcrowding
in the London area is a higher priority than putting extra vehicles
on this service. Between Coventry, Birmingham and Wolverhampton,
Centro is largely reliant on Virgin Trains to carry many of the
local passengers, and their recent upgrade to Voyager rolling
stock has created significant difficulties due to the severe overcrowding
on their services.
LIGHT RAIL—EXPERIENCES
ON MANCHESTER
METROLINK
41. Increased patronage of the Metrolink
network has resulted in parts of the system suffering from temporary
overcrowding, in particular during peak times and before/after
major public events. The success of the Metrolink system in moving
large numbers of people quickly and safely is well proven. During
the recent Commonwealth Games an estimated 80% of all visitors
used public transport in preference to car to access the events.
42. There are currently 32 trams in total
in the Metrolink Fleet and three services are planned to operate
as double trams in the morning peak. The current operator of the
Metrolink system has been unable to consistently meet the timetabled
requirement for trams in daily service due to a higher than expected
level of faults and failures. This has at times randomly increased
the levels of overcrowding experienced.
43. GMPTE is working closely with the existing
Metrolink Concessionaire and shortlisted bidders for the Phase
3 extension to address both the immediate requirement to remove
overcrowding and to ensure that the system will provide sufficient
capacity for the expected future patronage growth.
44. To address immediate overcrowding, GMPTE
has supported the current operator in providing increased peak
fleet availability by moving tram maintenance to night times.
It is also sponsoring the resolution of all latent defects in
the trams including a major overhaul of the 26 vehicles in the
Phase 1 fleet. GMPTE is currently investigating the opportunity
to procure up to 13 additional trams, which are expected to become
available shortly in Germany. These would be funded by a grant
of £6 million from the DfT, which has already been approved
in principle. The second-hand trams will provide a vital stop-gap
before new trams are built and commissioned on to the Phase 3
network.
45. GMPTE believes that the current demand
to use the system is partially suppressed by the overcrowding
and that the new journey opportunities and system quality will
substantially increase patronage. It has, therefore, set out clear
requirements to the bidders for the Phase 3 concession to provide
sufficient vehicles for the expected increase in patronage of
the existing system as well as the new extensions for the next
20 years. GMPTE is delighted that the DfT has endorsed this strategy
within the £520 million of funding approved for the Phase
3 expansion.
CONCLUSIONS AND
RECOMMENDATIONS
46. PTE areas have witnessed some significant
growth in heavy rail patronage and rail is a vital element of
their integrated transport systems. Growth has not, in general,
been accompanied by any significant investment in rolling stock.
This coupled with poor service delivery has led to significant
overcrowding problems (to the extent of passengers frequently
being unable to board trains). If the problem continues, the growth
is likely to stagnate and targets are unlikely to be met.
47. Many trains in PTE areas currently operate
as two, three or four car formations as standard. Most peak trains
do not operate at their maximum length and there is therefore
considerable scope to lengthen trains in peak periods to provide
additional capacity. In some cases, platforms need to be lengthened,
but this can be undertaken at considerably less expense than the
massive infrastructure schemes that are necessary to provide additional
capacity in the South East.
48. PTEG suggests that the following measures
are required to address the problems:
— new franchises need to include a
clear mechanism for ensuring that the provision of additional
capacity is linked to patronage growth and delivered in a timely
manner;
— investment is focused on longer trains
and platform lengthenings to maximise use of existing route capacity;
— franchises need to have adequate
resilience built in (including a sufficient level of spare rolling
stock to deal with day to day perturbations);
— action is required to correct failures
in the rolling stock leasing markets including the inconsistency
of lease lengths and franchise lengths;
— a national rolling stock strategy
should be developed without delay that includes a strategic pool
of rolling stock that can be allocated in the short term to deal
with peaks in demand.
December 2002
| 
