Pre-Feasibility Study for Bus Rapid Transit Hyderabad, Andhra Pradesh The Institute for Transportation and Development Policy With financial support from: US Agency for International Development Draft Final March, 2005 1 Executive Summary Hyderabad urgently needs to develop a mass transit system protected from its increasing traffic congestion. Mass transit is needed now on three central corridors, and will eventually be needed on at least nine corridors. Bus rapid transit (BRT), metro, and monorail technologies could all provide a system with sufficient capacity and speed to improve mass transit ridership in those corridors. Metro and monorail would require large capital subsidies, and subsidies for the procurement of rolling stock. By contrast, BRT would not require subsidies for operation, construction or for the procurement of rolling stock. The capital cost differences are significant. For the same amount of capital investments, ITDP estimates Hyderabad could build 294 km of BRT system, but only 37 km of elevated Metro or 31 km of monorail. The primary advantage of the monorail system is that it would minimize land acquisition. Metro’s main advantage is its potentially higher capacity, but this level of capacity is unlikely to be required in Hyderabad. While all 3 systems would save travel time and so help retain passengers on public transit, the metro and monorail system would require higher fare prices which would suppress demand. In addition, a larger system has much higher potential to shift passengers to mass transit. We project that if Rs.5000 crore were spent on each system, the BRT system would lead to a 21% increase in public transit mode share. By comparison, we estimate the same amount spent on metro would lead to a 1% increase in modal share, and for monorail would keep modal share constant (a 0% increase). BRT can provide effective mass transit without requiring an increase in fare price. Thus, BRT has the highest potential to increase public transit mode share and relieve congestion in Hyderabad. A 24 km BRT system in Hyderabad could be built using: o Private build-operate-transfer (BOT) or government investment for infrastructure: Rs. 123 crore (Rs. 5.1 crore/km) o Bus operator purchase of rolling stock: Rs. 80 crore For an optimal mass transit system, the government needs to improve the right-of-way for both vehicles and pedestrians, at an estimated cost of Rs. 68 to 200 crore. 1.1 Overall Traffic Conditions in Hyderabad The quality of life and economic vitality of the Municipality of Hyderabad are seriously threatened by the rapid growth in polluting two-wheelers, cars, and auto rickshaws, which have been growing at 10% per annum. Hyderabad’s air pollution level is much higher than the recommended World Health Organization standard for suspended particulate matter, resulting in thousands of pre-mature deaths each year. Traffic flow is nearing capacity, so that any minor incident can bring traffic to a standstill for extended periods of time. The rapid growth in the private vehicle fleet means that if nothing is done, Hyderabad’s traffic condition will deteriorate further and result in severe congestion. Even if congestion causes only 5 minutes of additional travel time per trip in Hyderabad, this means Rs. 254 crore per year lost to congestion. Worldwide, cities that have not developed an effective mass transit system have been unable to reduce congestion despite massive expenditures on new roads. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 2 Hyderabad’s buses are currently overcrowded during peak hours. Good average bus speeds of 18-20 km/h during peak hours are achieved, but this is largely because bus stops are 800 meters apart on average (optimal distance is between 400 meters and 500 meters). This creates significant passenger inconvenience from increased walking times, and high safety risks to passengers who board between stops. In congested areas, bus speeds have dropped to 11 – 12 km/h. As buses are caught in worsening congestion, they face increasing operating costs per kilometer. APSRTC, one of the better public operators in India, is today losing some Rs. 41 crore per year, and these loses are increasing. While APSRTC should be commended for providing reasonable bus service despite losses, the inconvenience and unattractiveness of Hyderabad’s bus system encourages further shifts to private vehicles. Traffic safety and the quality of public space are poor. In 2004, there were 419 people killed in Hyderabad city, most of them pedestrians. Thirty-three fatalities occurred along proposed Corridor I. These figures are exceptionally high by international standards. Sidewalks are non- existent or fully obstructed in most places, forcing pedestrians to walk in the road, compromising both traffic flow and safety. 1.2 Weighing Mass Transit Options: Basis of Analysis Given these conditions, Hyderabad needs to explore immediately its options for improving access to the city center, and the public space and walking environment within the city center. The options available for addressing this problem are widely known. There are three main proposals being discussed for Hyderabad: o Elevated metro o Elevated monorail o Bus Rapid Transit (BRT) Whatever measures are taken to improve mass transit in Hyderabad, they should be accompanied by policies to: o Restrain private vehicle use o Improve conditions for cycling and walking. ITDP recommends a combination of these two policies and BRT. ITDP encourages Hyderabad to seek more detailed analysis of the feasibility of the alternative options. This report analyzes the feasibility of BRT on a major corridor in Hyderabad, providing a preliminary assessment of: a. Whether BRT can accommodate the existing and projected future demand for public transit at a speed significantly higher than projected speeds for the existing bus system. b. Whether BRT can operate at a profit in this corridor. c. Whether BRT will also improve livability, reduce air pollution, and reduce traffic fatalities and injuries, and increase land values in the corridor. BRT is a high quality, ultra-modern and passenger-oriented mass transit system that delivers fast, comfortable, economical and eco-friendly mobility to urban dwellers. The most important feature of the system is physically segregated lanes for buses, private motor vehicles and non- motorised traffic such as pedestrians and bicyclists. Segregated lanes help in increasing the average speed of all motor vehicles (including buses) and improving the overall traffic flow. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 3 Some of the main characteristics of BRT are: o Segregated busways o Rapid boarding and alighting o Priority at traffic signal points o Clean, secure and comfortable stations and terminals o Modal integration at stations and terminals o Smart ticketing and efficient pre-boarding fare collection o Effective licensing and regulatory regimes for bus operators o Attractive, high capacity and customer-friendly buses o Clear and prominent signage and real-time information display o Sophisticated marketing identity Some of the major cities where this system has been developed or is in the process of development include: Bogotá, Sao Paulo, Curitiba, Mexico City, Panama City, Quito, Boston, Eugene, Chicago, San Francisco, Vancouver, Leeds, Strasbourg, Bradford, Lyon, Jakarta, Beijing, Kunming, Taipei, Nagoya and Seoul. 1.3 Estimating Future Demand for BRT, Metro, and Monorail ITDP concurs with the judgment of the Municipal Corporation of Hyderabad that the following three corridors have the highest potential demand for any mass transit option: o Corridor I. Dilsukhnagar to Kukatpalli o Corridor II. Secunderabad Station to Charminar o Corridor III. Mehdipatnam to Lakdi ka Pul We agree with the Municipal Corporation of Hyderabad (MCH) that Corridor I is a rational first priority both because of its high demand and the reasonable availability of existing right-of-way. 1.3.1 Existing transit demand in Corridor I Any analysis of future demand must begin with an estimate of existing demand in the corridors. According to the Municipal Corporation of Hyderabad, Corridor I currently carries 40,000 to 50,000 peak period passengers per hour per direction (pphpd), of which 34% of the trips are carried by bus. Thus, the current number of bus passengers in the corridor is as high as 17,000 pphpd. Our measurements agree with this estimate. The existing maximum load on the most critical link (static demand) is 11,600 pphpd. Adding the passenger rotation ratio (accounting for passengers getting on and off) of about 1.5, the total passenger trips one way is 17,400 pphpd. Daily demand on Corridor I is around 307,000 passengers per day. A system able to capture the entire existing demand on Corridors I, II, and III, would have a total daily demand of around 686,000 +/- 20%. We recommend a 300,000 On-Board Origin- Destination Survey of existing bus and paratransit passengers be carried out as a first step in project development, to allow more accurate estimation of demand. 1.3.2 Projecting Future Modal Shift Future demand projections should start with historical trends. Bus trips have remained fairly steady at around 3 million trips per day since 1996, falling slightly in recent years to around 2.8 million trips per day (APSRTC, 2004). Therefore, if no new mass transit system is built, we can Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 4 expect public transit ridership to remain constant or face modest declines. As population continues to grow, this means that transit would lose mode share if nothing is done. How many passengers can be induced to switch from other modes to transit if a new mass transit system is built will depend on the characteristics of the system built. Projected future congestion levels for mixed traffic, and the level of restriction on private vehicle use will also affect modal shift. Because metro and monorail systems are expensive, they generally can not be expanded fast enough to keep pace with urban dispersal and the growth of private motor vehicle fleets. No metro system in the world for which data is available has led to an aggregate modal shift in favor of public transit away from private motor vehicle use. Therefore, it would be wrong to assume that a metro or monorail in Hyderabad would have a significant modal shift impact. Public Transit Modal Split Before and After BRT and Metro Construction City % of Trips Before % of Trips After Metro Systems Mexico City 80 72 Buenos Aires 49 33 Bangkok 39 35 Kuala Lumpur 34 19 Santiago 56 33 Warsaw 80 53 Sao Paulo 46 33 Tokyo 65 48 Seoul 81 63 BRT Systems Bogotá 53 56 Curitiba 74 76 Quito 76 77 Sources: OTP (2003), Xu, K. (2004), Vasconcellos, E. (2001), SETRAVI (2003), Ciudad Viva (2003), WBCSD 2001, Barter 1999, Kuala Lumpur Draft Master Plan 2003, National Transport Secretariat of Argentina 2001, TransMilenio SA, Bogotá, 2002. URBS, Curitiba, 2004; BRT, by contrast, has proven to bring about a modest modal shift in favor of public transit trips, primarily because fares can be held lower and the system can grow to reach a lot more people for the same investment and implementation period. To project actual ridership for a new mass transit system in Hyderabad, it is not safe to assume that even all of the current transit passengers in the corridor will use the new system. The factors which determine the percentage of total transit trips that will actually be captured by a new public transit system serving this corridor are: o The number of trips which both originate and end along the corridor. o Whether or not normal bus routes are allowed to continue on the corridor o The fare price of the new mass transit service relative to any competing mode choices available in the corridor o The door to door travel time of trips utilizing the new mass transit service (inclusive of transfer time) relative to other modal choice options in the corridor. o The feeder system provided and its cost. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 5 Currently, only 40% of the passengers currently using Corridor I are both beginning and ending their trip directly along that corridor. Many existing bus lines take Corridor I to other destinations. If all existing bus lines are allowed to continue to use Corridor I in competition with any new mass transit system, the maximum load on the critical link (static demand) will be only around 7000 pphpd on the new mass transit system. As the presence of the old buses in the mixed traffic lanes would not only undermine the profitability of the new system but also congest the mixed traffic lanes, it is recommended that whatever mass transit system is designed for the corridor, the following measures be taken: o 80% of competing bus lines in Corridor I should be cut. o Free integration with mass transit service or bus services in trunk Corridor II and III should be provided. o Feeder buses should be provided at the terminals of Corridor I and at some intermediate points 1.3.3 Projected Demand for Specific Mass Transit Options Estimating the likely future demand of the monorail, elevated metro, and our BRT proposal, requires defining very carefully the characteristics of these systems. Unlike with the existing bus system, we can safely assume that because a metro, a monorail, or the BRT system would all have dedicated rights of way, increasing congestion of general traffic will improve the comparative advantages of traveling by mass transit. Therefore, we have assumed that if any of the three mass transit systems are built, transit ridership will increase slightly faster than population growth in the specific corridor it serves. However, the systems have different characteristics that will have a profound impact on the degree to which they can attract passengers. This report tests three specific scenarios: 1) the elevated metro proposal made by DMRC, 2) a theoretical monorail proposal based on a similar proposal made in Jakarta, and 3) a BRT system with three possible mechanisms for system integration with Corridors II and III. Based on the system characteristics from the pre-feasibility studies, we estimate the following projected future demand for each system: Comparative Demand for Monorail, Metro, and BRT BRT Metro Monorail System Corr I System Corr I System Corr I Daily Pax pphpd Daily Pax pphpd Daily Pax pphpd 2008 854,001 14,441 653,862 11,057 482,362 8,157 2011 905,221 15,307 693,093 11,720 511,303 8,646 2021 1,076,042 18,195 873,298 14,767 607,776 10,277 1.3.3.1 Characteristics of the DMRC Proposed Elevated Metro Rail System The DMRC proposes a 38.3km elevated metro built on Corridor I (25.6 km connecting Miyapur and Chaitanya Puri) and Corridor II (12.7km connecting Secunderabad railway station and Falaknuma Railway Station). The metro being proposed by the DMRC would be elevated. The proposal saves money and increases system speed by constructing fewer stations, but this will reduce convenience and potential modal shift to the system. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 6 The elements of the DMRC proposal that are likely to affect demand in the corridor are: o Transit vehicle speeds would increase from an existing average speed of 18 – 20 km/h to some 34 km/h. o The average metro fare would be 50% higher than the projected bus fare. o The 1 km distance between stops would be 25% farther than the existing 800m between bus stops. 1.3.3.2 Characteristics of the Metrail/Nash Proposed Monorail System We understand that MetRail of Switzerland and Fraser Nash of Great Britain have both proposed monorail projects for Hyderabad. Having no detailed proposal, we consider a likely scenario: o A 55 km monorail line would be built on Corridor I and II by private investors in a Build-Operate-Transfer (BOT) scheme. o Based on Kuala Lumpur experience, the speed of monorail system could be as high as 30 km/h, depending on the distance between stations, attracting ridership. o Fares are more than 50% higher than projected bus fares at time of completion. o Stations are in roughly the same locations as proposed by the metro company. 1.3.3.3 Characteristics of ITDP’s proposed BRT System The system characteristics for BRT are detailed in the following section. However, for demand estimation purposes, our proposed BRT system will have the following basic characteristics: o A 24km long system o Average speeds of 26kph o 450 meters between station stops. o Fare prices equal to the projected future bus fares. Four methods of capturing the majority of existing and future transit demand in Corridor I using BRT are presented, one of which was rejected. The first three BRT options would all build a ‘closed’, Curitiba or Bogotá-style, BRT system down Corridor I. The first option would include two transfer stations to be built in the city center that would allow free transfer for passengers traveling from existing buses operating in mixed traffic on Corridor II and Corridor III onto the trunk mass transit line in Corridor I. This scenario would ensure the demand on Corridors II and III onto Corridor I would be captured. The second option utilizes the ‘closed’ BRT system on Corridor I, and for Corridor II and III would procure buses that can operate both on the BRT system and also in mixed traffic. Operating both on and off the BRT system, these buses would be able to bring most of the demand onto the BRT system in Corridor I. The third option provides passengers with free transfers between the normal bus system and the BRT system anywhere along the corridor through use of a smart card ticketing system. This option has the advantage that it would have the maximum impact on demand. The main disadvantage is that it requires the procurement and installment of a smart card ticketing system on all buses in Hyderabad. Any one of these first three, ‘closed’ BRT, options will satisfy the stated project goals. The demand estimates for all three of these scenarios will be similar and, within the limits of our analysis, the costs similar. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 7 A fourth BRT design option was considered and rejected. This option was to design an ‘open’ BRT system using the existing normal buses but giving them exclusive lanes in the center of the carriageway, as is being developed in Delhi. This option would reduce transfers and would ensure that all of the transit demand in the corridor could be captured. Because an open busway would either have very slow operating speeds or badly congest the mixed traffic lanes, or would require a lot more land acquisition, we do not recommend the option of designing a Delhi-style ‘open’ BRT system. 1.3.4 Effect of Fare and Travel Time on Demand for Three Alternative Systems The different systems being proposed will have different impacts on potential ridership. For comparison, we have assumed that all three systems will cut bus lines in Corridor I and capture demand from Corridors II and III through some sort of free transfers. Thus, the main factors varying demand between the three systems are differences in door-to-door travel time and door- to-door travel costs. Door-to-Door Travel Time for Three Mass Transit Systems Metro Monorail BRT Speed Dist- Minutes Speed Dist- Minutes Speed Dist- Minutes (km/h) ance (km/h) ance (km/h) ance (km) (km) (km) Walking 4 0.5 7.5 4 0.5 7.5 4 0.25 3.8 Waiting - - 3 - - 3 - - 2 Riding 34 9.2 16 34 9.2 16 26 9.2 21 Walking 4 0.5 7.5 4 0.5 7.5 4 0.25 3.8 Total 34 34 31 Time An analysis of door-to-door travel time is presented in the figure above. The estimates for walking time for the metro and monorail are conservative; they include no factor for the increased time and inconvenience of having to climb the flights of stairs necessary to reach the elevated stations. For comparison, a similar trip on Hyderabad’s current bus system would have a door-to-door trip time of 51 minutes. Applying the value of time and cost elasticity assumptions used by DMRC (time valued at Rs.10/hour and cost elasticity =0.5) to all three scenarios, we can predict the effect of the travel time savings on the baseline demand. Travel Time Effect on Demand for Three Mass Transit Systems Metro Monorail BRT Time Savings (minutes) 17 17 21 Equivalent Value (rs) 2.8 2.8 3.4 Percentage savings 41% 41% 49% Effect on Demand +20% +20% +24% Using the DMRC’s assumptions for projected fares for regular buses and the metro, plus an estimate of fares based on the policies adopted by other BOT monorail projects in the region, we can estimate the effect of cost on demand. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 8 Travel Cost Effect on Demand for Three Mass Transit Systems Metro Monorail BRT Projected Fare 10.5 14 7 Percentage difference from bus 50% 100% 0% Effect On Demand -25% -50% 0% The net effect of the differences in system characteristics between the three mass transit systems analyzed on demand is -5% for metro, -30% for monorail, and a +24% for BRT. 1.4 System Design Recommendations Metro, BRT, and monorail could all handle the projected passenger demand in the corridor. The DMRC metro proposal would have an initial capacity of 20,000 pphpd on corridor I and 12,000 pphpd in Corridor II at the beginning of operations, increasing to 49,632 pphpd in Corridor I and 31,020 in Corridor II by 2021. Given a projected initial demand estimate of 11,000 pphpd, we believe this system will provide more capacity than needed. A monorail system in Corridor I could carry about 18,000 pphpd if it were four cars long, requiring the elevated stations to accommodate four car trains. Metrail and Frazer Nash are claiming 36,000 pphpd, but this has not been achieved by an existing monorail. Given our projected demand estimates, we believe the monorail would be more profitable if it were designed to carry only around 10,000 pphpd. This is because the higher fares of monorail will suppress the demand to this level. We recommend designing a BRT system to handle 18,000 pphpd upon opening in 2008 and increasing its capacity to 36,000 by 2021. While this is more than the projected demand in 2008, the additional cost of designing a system to handle this level of demand over projected demand is marginal. The metro and the monorail systems can be designed with sufficient capacity to handle projected future demand. This report explains how these capacities can be achieved using BRT technology. A BRT system can be built in Corridor I with the nearly the same capacity as the proposed metro system, with an average operating speed of 26 km/h, starting at 18,000 pphpd and increasing to 36,000 pphpd over time. TransMilenio in Bogotá is transporting 38,000 pphpd. Reaching these levels of capacity in a BRT system requires careful design and engineering, utilizing the following characteristics: o The BRT system should occupy the central verge of the roadway, rather than the curb lanes. This will avoid conflicts with turning traffic, pedestrians, stopping taxis and delivery vehicles, illegally parked vehicles, etc. o Passengers should pre-pay to enter each bus station, and each station platform should be elevated to the height of the bus floor. o The exclusive bus lanes must be physically separated from the rest of the traffic by a physical barrier, and enforcement of encroachment onto the busway must be maintained with additional police at the intersections during the initial months of operations. Fines for illegal encroachment on the busway must be strictly enforced. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 9 o An overtaking lane at stations. The busway needs two lanes in each direction at each station, and one lane in each direction at all other points. The overtaking lane is critical to relieve bus congestion. o Each station should have at least two platforms. o The distance between bus stations should be brought down from the current average 800 meters to an optimal level of 450 meters. This will slow down bus speeds somewhat (from 30 to 26 km/h) but it will reduce total trip time by reducing walking distances. Forty stations with 80 pre-paid enclosed platforms should be built along Corridor I. o For Corridor I, use 109 articulated buses with 150-passenger capacity and four platform-level 1.1 meter wide doors on the right side (Scenario I and III), or the above buses plus 206 buses with two platform-level 1.1 meter wide doors on the right side and two standard curb-level left side doors on the left (Scenario II). o Pedestrian access to the central verge should be at grade using improved crosswalks rather than using pedestrian overpasses. As the maximum number of mixed traffic lanes the pedestrians would need to cross is two, this can generally be negotiated safely. (ITDP recommends pedestrian overpasses only when three or more mixed traffic lanes need to be crossed, with average vehicle operating speeds over 40 km/h.) Even in presence of overpasses, people generally prefer to negotiate the crossing at grade. o Where possible, restructure right turns along the corridor to increase bus speeds and avoid worsening congestion in mixed traffic lanes. The capacity of the system will be expanded from 18,000 pphpd to 36,000 pphpd as needed by adding an additional bus platform at each station, and by adding more express bus services which make fewer stops. The addition of this extra BRT capacity costs very little money, in contrast to the cost of expanding the capacity of a metro or monorail system. An advantage of BRT systems over rail based systems is that headways between each vehicle can be reduced from around 3 minutes to less than 30 seconds. Because the obstacle to reducing headways is the capacity of the bus stations, additional capacity is added to a BRT system by adding additional bus platforms at each station. For this to work, however, an overtaking lane at the station is necessary. 1.5 Right-of-way The right-of-way in Hyderabad’s Corridor I is highly irregular, as is typical of many older cities, varying from over 60 meters in width to as little as 9 or 10 meters at specific bottlenecks. Any mass transit system proposed for this corridor will have to address the irregularity of this right- of-way. This Corridor, most of which falls along National Highway 9 under the authority of the National Highway Authority of India (NHAI) is slated to be upgraded/widened from the existing 4 lanes to 6 and 8 lane roads. Efforts for BRT in Hyderabad therefore should be coordinated with the NHAI widening. Throughout most of the right-of-way, walkways and cycle paths are either non-existent or highly inadequate. This compromises the quality of the urban environment in the city center, making this an unattractive and highly inconvenient destination for shoppers or employers. Regardless of the type of mass transit system selected for this corridor, the MCH needs to significantly improve the quality of the walking and cycling facilities in the corridor. Hyderabad BRT Pre-Feasibility Study, Draft Final ITDP – March 2005 10