The Metro Express: some inconvenient truths analysed
4 Mai 2017 - lexpress.mu
In this series of articles, the authors try to build a comprehensive picture of the economic, infrastructural and social problems related to the Metro Express project.
They show their preference to a bus route rather than light rail.
The Metro Express: some inconvenient truths analysed (Part I)
The Metro Express (ME) project continues to generate debates in both the National Assembly and the wider public. This is highly desirable in view of its immense costs and questionable technical specifications. Our millennials and their successors will bear the heavy financial burden and it is desirable that they join the fray.
There is clearly a strong desire on the part of the current government to introduce the ME. The Singapore Cooperation Enterprise (SCE) submitted its first report on the introductionof the Light Rail Transit (LRT) to the former government. Following a Private Notice Question (PNQ) by the then Leader of the Opposition on 22 July 2014, the then Minister responsible for transport informed the House the LRT project would cost Rs 22.2 billion and tabled an executive summary in the National Assembly, to which everyone is referring today.
In developing the proposed LRT, it would be interesting to know if the SCE has carried out detailed feasibility studies on the following issues:
- Patronage on existing bus services and surveys of bus users;
- Analysis of passenger car data;
- Surveys of car users and the likelihood to transfer to the LRT service;
- Assessment of existing and future traffic congestion;
- Assessment of impact of socio-economic land use changes;
- Revenue analysis, financial and economic;
- Technical analysis of preferred route and engineering solutions;
- Detailed route assessment and description of preferred alignment;
- Civil engineering assessment of structures;
- Electrical assessment of power supply;
- Rolling stock assessment of LRT solutions;
- Traffic management assessment of LRT/vehicle interface;
- Institutional and contractual assessment;
- Environmental assessment.
The aim of these feasibility studies would be to develop the proposed ME network technically and economically into a “bankable” proposal. This can then be used to seek funding from local/international organisations and would have guided contracting organisations wishing to design, construct and/or operate theproposed network.
If detailed studies have been carried, then they should be made available to the public. If not, this is a serious flaw. The public must know in advance what kind of financial situation it is getting into before deciding on a project. A major reason for the financial failure of the first two LRT lines in Kuala Lumpur (KL), in Malaysia, and the monorail project was the poor and inaccurate information given to the government, preventing it from making the correct assessment.
According to a study by Jeff Tan (2008), the KL system was privatized through a build-operate-own concession, based on grossly inflated passenger projections presumably to justify financial viability to secure financing. The ridership was way below projections and insufficient to cover operating costs, let alone interest and principal repayment.
In a communiqué dated 14 April 2017, our Ministry of Public Infrastructure and Land Transport informed the public that the total expenditure incurred as at to-date amounts to Rs 572,239,337.40. From the break-down of the figures published, Rs 1,15 million were paid to Grant Thornton for household and traffic surveys and Rs 5,843 million to Water Research Co. Ltd for geotechnical survey. It would hence appear that, apart from the traffic surveys, none of the above studies has been carried out.
Generally, good ridership numbers are required in order for operating revenue to break even with the costs. In response to the PNQ from the Leader of the Opposition on 28 March 2017, the Honourable Bodha, Minister of Public Infrastructure and Land Transport, announced that the ME is anticipated to carry 80,000 passengers per day per direction. There have been no details explaining how this figure of 160,000 has been arrived at.
Put into context, the three bus operators – NTC, UBS and RHT – carry approximately 160,000 passengers daily on some 400 buses across 35 bus routes from the Plaines-Wilhems to Port-Louis and vice-versa. Buses will continue to operate along their existing routes. No question about that.
If, say, 50% – which is a high estimate by the way – of the bus passengers shift to the ME, then the total number of passengers on board the ME will be 80,000 for both directions giving an average flow of nearly 2,700 passengers per hour (pph) per direction over a 15-hour period. To make up for the other 80,000 passengers, the occupants of at least 25,000 cars – assuming 1,5 occupants by car – would have to be shifted!
In his article “Are politics and prestige trumping sound finance and credible economics? (Part II)” – lexpress. mu, of 17 March 2017 – Dr Rama Sithanen stated that “previous studies show that ME is viable onlyif there are at the very least 15,000 passengers per hour, which is impossible along the Curepipe–Port-Louis corridor.” As a matter of fact, the number should be at least 10,000-12,000 pph per direction over a period of at least 12 hours to make the ME break-even, representing an average of 264,000 passenger trips per day, a figure which is more than impossible to attain!
It has been mentioned in the defimedia.info of 3 March 2017 that 30 trains would be sufficient to service the corridor between Curepipe and Port-Louis with a headway (i.e. the time interval between two trains) of 3 minutes. It is assumed that the trains will each have a full carrying capacity of some 400 passengers. With a 3-minute headway, the ME would have a capacity of 8,000 pph per direction during the peak periods. These are small numbers and could easily be carried by a Bus Rapid Transit system, the more so as the flow may only be 3,000 pph per direction if there is some sort of shift from the private cars to the ME.
Mauritius simply does not have the necessary critical mass to support the ME. Reference has been made by many to Singapore, Australia, Hong Kong, etc. People tend to forget that likes should be compared with likes. The Singaporean rail operator SMRT Corporation Ltd and regulator Land Transport Authority are deciding what to do with the problematic Bukit Panjang LRT, including scrapping the system altogether (Straits Times, 11 October 2016).
In a company blog, the SMRT Trains Managing Director, Lee Ling Wee, said that the 17-year-old system is near “the end of its design life”. He added that an idea to do away with the entire LRT system was also mooted and for residents in the Bukit Panjang area to go back to riding buses. He added that “this is not far-fetched, as a fully loaded high-capacity bus like a double-decker can take 130 passengers, which is more than the 105-person capacity of a single Bombardier train.” He reckoned however that this option would of course lead to more road congestion.
A group of Transport Hobbyists
The Metro Express: some inconvenient truths analysed (Part II)
The Gold Coast LRT
Mr Georges Chung, Senior Adviser at the Prime Minister's Office and Director of the Metro Express (ME) Ltd, in a statement to Radio Plus on 01 April 2017 stated that the Gold Coast Light Rail Transit (LRT), in Queensland, Australia – to which he makes reference so often – has been implemented for a population of some 530,000. What he forgot to say was that Gold Coast has an influx of 12 million tourists annually and 12,500 new residents each year.
In an interview to Defi Media (defimedia.info, April 8,2017), Mr Chung's reply to a question from the journalist regarding the journey time by the ME was that "à Gold Coast, en Australie, il y a 16 stations et le temps couvert est de 33 minutes". He forgot to say that the time was for a distance of 13 km only! This means that the speed of the Gold Coast LRT is 23.6 km/h.
If this speed is applied to the 26 km distance in Mauritius, the journey between Curepipe and Port-Louis would take 68 minutes. When the time for stopping (say 20 seconds) at each of the 17 stations between the origin and destination is included, the total journey time would be 74 minutes as compared to the 43 minutes stated by Mr Chung.
In order to perform the journey in 43 minutes, the ME should travel at a speed of 37 km/h, which is quite too fast when one looks at the average speed of 25-30 km/h of LRTs worldwide. One should not forget what happened in November last when the LRT between Wimbledon and Croydon (London) exceeded its average speed. It resulted in a serious accident with seven deaths.
For the information of the public, the following is submitted as regards the Gold Coast LRT -
- The distance served by the LRT is 13 km long with 16 stations.
- The cost of the project was nearly 1.3 billion Australian Dollars (AU$), equivalent to Rs 36 billion (Rs 2.769 billion/km) as compared to its projected cost of AU$ 949 million (Rs 26 billion).
- Construction works started in August 2012 and operation began in July 2014.
- Although it has a top speed of 70 km/h, its average speed is 23 km/h.
- The system consists of 14 flexy 2 trains, each 43.5 metres long and the 2 trains carry 309 passengers with seating accommodation for 80.
- The headway (i.e. The time interval between two trains) varies between 7 minutes during peak hours and 30 minutes for the off-peak periods.
- A second part of the LRT is an extension of 7.3 km, with two intermediate stops and the journey time will be 11 minutes. Works started in July 2016 and are expected to be completed in April 2018. The cost of the extension is estimated at AU$ 420 million (Rs 12 billion) (Rs 1.643 billion/km).
- Estimated population as at 30 June 2016 - 567,644.
For the ME to utilize its ultimate capacity of 8,000 passengers per hour, there must be sufficient feeder bus access to the different stations, enough large park and ride spaces adjacent the stations, and/or enough people within walking distance of stations. Shortcomings in any of these areas will limit the useable capacity of the ME.
The financial viability of ME would be critically dependent on its ridership and fare structure; otherwise, it may end up as a white elephant waiting to be bailed out by the public. Will the projected ridership of 160,000 ever materialise? If it does not – which is most unlikely as stated above – how will the country be able to support such a costly project as the Metro Express?
Despite the recent commitments to capital funding for the proposed ME, there is no clear idea of how much each part of the ME project will cost in terms of construction, much less the annual cost of operating and maintaining (O&M) the proposed system. There has been no discussion on how the system's operating costs will be paid for after it starts operating. Any funding commitment to building the ME means nothing if the public doesn't know how it is going to pay for it after its construction. It is a known factthat it is the O&M costs that will make or break the financial viability of the project.
Lack of transparency
As the Government is supposedly committed to good governance
- accountability and transparency amongst others
- there is a lack of clarity with regard to the financing of the entire project. We know that the Indian government has given a grant of some Rs 9 billion to the Mauritian government for the ME.
- Will the difference between the grant of Rs 9 billion and the final cost of the ME be financed by loans from commercial banks, guaranteed by government and for equity participation by local/foreign investors?
- What is the public being committed to?
- What happens if funds run out after the initial stages?
- Will the public be subjected to unaffordable fares, thereby making them return to their cars?
Government-guaranteed loans will add to public debt and the private sector investment will look for the internal rate of return of the project. Nothing is clear at this stage. Questions are being asked, but very few answers are being provided.
A group of Transport Hobbyists
The Metro Express: some inconvenient truths analysed (Part III)
Construction timelines and costs
The construction time for a Bus Rapid Transit (BRT) is significantly shorter than that for the Light Rail Transit (LRT). Measured in months taken to build one kilometre for different cities, the BRT takes an average of 1.4 months per km to construct, with a range of between 0.7 months per km in Yichang, China, and 2.3 months per km in Caen, France. As every system is different, it is difficult to compare construction timelines of different projects. However, on average, the construction of LRT appears much longer, with an average of 2.8 months per km
To construct a BRT at grade over the same distance, based on the average of 1.4 km per month from the table above, it would take just 36 months, which is almost twice as quick as for the LRT. In the case of the LRT, significant disruption and added traffic congestion will be caused by transport infrastructure construction and will result in frustration for commuters, businesses, residents, schools and other institutions along the alignment. Compared to LRT, the construction of a BRT would be far quicker and disruptions would be minimised
The construction cost per kilometre for the Honolulu LRT is 3.4 times that of the three other cities. Excluding the Honolulu LRT, the average construction cost per kilometre is US$ 62 million (Rs 2.26 billion) as compared to US$ 16 million (Rs 582 million) for the BRT. Applying this cost of US$ 62 million per kilometre to the 26 kilometres of the Metro Express (ME), the cost amounts to US$ 1.6 billion or Rs 58 billion at the current rate of the US$ against the Rupee (Rs 36.40). If the average cost of Rs 2.206 billion per km of the Gold Coast and its extension is taken as reference, then the 26 km of the ME will cost Rs 57 billion.
The most recent LRT project is the LRT 3 line, connecting Bandar Utama and Klang in Malaysia (2017). The line is 37 kilometres long with 26 stations and will cost Rs 76 billion (RM 9 billion @ Rs 8.50) excluding land reclamation costs. The system will be elevated over 35 kilometres.
The ME was estimated at US$ 850 m (Rs 31 billion at today's rate) but its cost has been brought down to Rs 17.7 billion (55% of the initial cost) as it seems that the number of stations has been reduced by 2 and the system will not be elevated throughout but will operate mostly at grade level. Taking the LRT 3 as reference, the ME will cost at least Rs 45 billion.
It is recognised that the construction costs in table 2 relate to foreign systems and that the cost rates in Mauritius are likely to be somewhat lower. Offsetting this, however, is that there is no rail manufacturing capability on the island and that all the equipment will have to be imported, and much of it will need to be installed by foreign engineers.
The estimates in working the operating costs of the ME are based on the construction of the ME running from Curepipe to Port-Louis, over a distance of 26 km.
For the purposes of costing requirements,
- the service is assumed to run:
- from 05h00 to 22h00 during weekdays and Saturdays (17 hours over 300 days);
- from 06h00 to 21h00 on Sundays and Public Holidays (15 hours over 65 days).
- Peak hours are assumed to be from 06h00 to 08h30 and 15h30 to 18h00 (5 hours for 300 days).
- The frequency will be:
- one train in each direction every 3 minutes during the peak hours;
- one train in each direction every 6 minutes during the offpeak hours;
- one train in each direction every 10 minutes on Sundays and Public Holidays.
- Assuming the average speed of the LRT to be 30 km/h, which is slightly above the international standard of 23- 25 km/h, the journey time will be 52 minutes if it is a non-stop service from end to end. If the ME stops for a minimum of 25 seconds at each of the 17 stations (excluding the stations for departure and arrival) forpassengers to board and alight, the overall journey time would be around 60 minutes. In addition to the journey time, a minimum turnaround time of 5 minutes is used. The combination of frequency and running time requirements generates a need for at least 42 trains to operate the service. This does not include any built-in spare provision. Provision should be made for four additional vehicles, one as operational spare and three as maintenance spares.
Estimated Costs and Costing Assumptions
Based on the above, the annual kilometres run would be about 3.7 million kilometres. This has been used to calculate any costs driven by distance.
The cost of energy constitutes a significant proportion of the overall operating costs as is the case with all modes of transport. To meet the gradient requirements from Port-Louis to Curepipe, the ME would operate most probably from 750-800 volts direct current.
Since details of the optimum power requirement is not known at this point, it is assumed that the energy consumption of the ME would be 10 KWh/trainkm for traction purposes and that for auxiliary items such as lighting, traffic signalisation, maintenance workshop, depot, etc., is normally about 20% of the traction energy.
Electrical power costs have been estimated using the Central Electricity Board preferential tariff for industrial users of Rs 5.40 per kWh. Based on the above, the total cost of energy would be Rs 250 million, inclusive of the monthly subscription.
A group of Transport Hobbyists
The Metro Express: some inconvenient truths analysed (Final part)
Estimated Costs and Costing Assumptions
The energy consumption of the Metro Express (ME) would create a significant additional power demand. As it seems that the power generation facilities are operating close to their maximum capacity, the CEB will therefore have to cater for additional power demand during the peak periods of operation of the ME.
Staff will most likely comprise -
- 66 drivers. This is derived from the number of 8 hour instances required to run the service 365 days a year.
- 10 inspectors and 8 supervisors. This allows for 3 shift cover plus rest day and general purpose relief cover.
- 57 station staff.
- 42 for central and line control.
- 80 for infrastructure maintenance.
- 45 for rolling stock maintenance.
- 35 for administration purposes.
- 12 cleaners.
Staff totals 355 and is assumed to work an average 40 hour week.
Annual total staff cost - Rs 75 million.
Maintenance costs – Rs 90 million
(Material costs for vehicle and infrastructure maintenance are based on Light Rail Transit (LRT) experience in other countries).
Depreciation cost of rolling stock, equipment and infrastructure - Rs 530 million.
Administration – Rs 95 million.
Total annual operating costs – Rs 1.04 billion.
Operating cost/train-km would be Rs 280 million.
Based on 3,000 passengers per hour per direction, over 15 hours, and a weighted fare of Rs 24.00 (based on actual bus fares), the annual fare revenue will amount to Rs 788 million.
The figure of 355 employees is 20 times lower than the figure of 7,000 put forward by Mr. Sudesh Lallchand, Senior Adviser to Minister Bodha, who stated in an interview to l'express that «des pourparlers sont en cours entre le gouvernement et les trois compagnies d'autobus pour qu'il n'y ait pas de licenciement.» He added that «de plus, 7 000 emplois additionnels seront créés quand le Metro Express sera en service.» (lexpress.mu - 28March 2017). He further stated in that same interview that, if he thinks that «le projet atteindra une somme astronomique, il n'y aura pas de Metro Express».
Prioritise public transport
Priority should be given to building a robust public transport system over road expansion. The latter only serves to undermine the former. For a public transport system to succeed, both the pull and the push factors are needed. One without the other will not do.
The pull factors are to provide convenient, punctual, frequent, affordable, accessible and well-connected public transport system. Government must have the political will to implement push factor policy measures to discourage the use of private vehicles.
In the short to medium term, action will need to be taken to restrict parking spaces, expand street parking controls and parking charges will need to be significantly increased; while in the longer term, road congestion charges, as recommended by Mr Gopinath Menon in his report Congestion Pricing in Port Louis, June 2004, need to be considered for vehicles entering Port Louis and other towns like Curepipe, Quatre Bornes and Rose Hill during the morning and afternoon peak periods.
Imposing charges on private vehicle users is a well-tried market-pricing mechanism. In fact, the charges imposed on people who still prefer to use private vehicles can be transferred to subsidise public transport travel. Contrary to the argument that such a policy is undemocratic and elitist, the opposite is true. The richand those who choose to use private vehicles contribute more to pollution and carbon emissions than the masses who take public transportation.
Imagine, a rapid transit system carrying hundreds of passengers will remove hundreds of single-occupancy vehicles from the road, reducing carbon emissions. Private vehicle users generate higher external costs for society and should pay for their actions. This is eminently democratic and fair. Climate change is upon us and Government should be setting its own carbon emission targets and streamlining development and transport policies to achieve them.
Having, say, a 60% public transport modal share will not happen by building public transport alone. Future growth of vehicle traffic should be limited. New road-building should be extremely selective and cautious to improve local connectivity and feeder routes but should not compete with public transport priorities. The modal share is around 40%, which is low, indicating a preference for private transport. For comparison purposes, the figure for Hong Kong is 80%, Tokyo 65%, Singapore 67% and London 52%.
Transport planning does not privilege car owners/users, but prioritises public transport that is more comprehensivein coverage and more accessible to a wider population, including the low-income, the elderly and people with disabilities. The success of a public transport system therefore hinges on its level of accessibility, connectivity and integration, and as far as possible, there should be a single integrated transport system.
Evidence-based policy and far-sighted leadership
Good policies should be evidencebased and analytically, not politically, driven. The crux of the problem in Mauritius, like many small cities, is that there are too many cars, with inadequate public transportation, competing for scarce resources. Improvement of public transportation must be accompanied by proven means to decrease the number of low occupancy vehicles, increase high occupancy vehicles and public ridership. Such a strategy requires our leadership to adopt a new paradigm and then to communicate and educate our public so they understand that this is the only possible way out.
Annual fare revenue, conservatively estimated at Rs 788 million, would only partly cover the annual operation and maintenance costs, estimated at approximately Rs 1.04 billion. The balance would therefore require an ongoing government subsidy. It is worth recalling that the first two LRT lines in Kuala Lumpur (KL) had to be bailed out at great cost by the Malaysian government in less than five years, due tolower-than-expected passenger ridership, although the operator had a 60-year concession. Similarly, the KL Monorail had to be taken by the government in 2007, four years after it began its operations, once again due to low passenger volume.
As the number of passengers per hour per direction would not meet the required number for the ME to breakeven, the best option for Government would be to opt for Bus Rapid Transit, which is being implemented in many regions as shown in the table below.
A group of Transport Hobbyists