Reducing Subway Overcrowding at the Manhattan CBD Cordons - Vol. 1: The Queens Cordon
Click HERE for graphic. Long Range Planning: Reducing Subway overcrowding at the Manhattan CBD Cordons PT2209893 PT220980L Contract D000642 Task 3.0 The preparation of this report was financed in part with funds from the U.S. Department of Transportation, Federal Highway Administration, under the Federal Highway Act of 1956, as amended, and the Urban Mass Transportation Act of 1964, as amended. This document it disseminated by the New York City Department of Transportation in the interest of information exchange. It contains proposals offered for discussion purposes by the New York City Department of Transportation, which is responsible for the facts and the accuracy of the data presented. The report does not necessarily reflect any official views or policies of the Federal Transit Administration, the Federal Highway Administration, the State of New York, or the City of New York. The report does not constitute a standard, a specification, or a regulation. Prepared by: New York City Department of Transportation Michael A. Weiss Chief of Staff Edward S. Seeley Jr. Deputy Assistant Commissioner Ann Marie Sledge, Director of Strategic Plan Development Deborah J. Molina, Erica Caraway, Graphics T A B L E O F C O N T E N T S I. EXECUTIVE SUMMARY 3 II. CURRENT CONDITIONS 25 III. HISTORICAL BACKGROUND 37 IV. SOLUTION OPTIONS 44 V. CONCLUSIONS 62 I. E X E C U T I V E S U M M A R Y This is the first of three reports discussing the issue of subway overcrowding at the cordon crossings of Manhattan's Central Business District (CBD) . This report focuses on the Queens cordon. The other two will focus on the 60th Street cordon and the Brooklyn cordon. (The New Jersey cordon is not dealt with in these reports, since no subway lines cross it. However, the public transportation needs at this cordon are being evaluated by the Port Authority's Access To The Region's Core study.) Each report will follow the same format and have five sections. 1. An executive summary. 2. A description of current subway conditions at the cordon. 3. Some historical background on the development of subway lines serving the cordon. 4. A description of various options for increasing subway capacity to eliminate overcrowding. 5. An outline of the report's conclusions. The report uses technical data that was provided by the New York City Transit Authority. An earlier draft was reviewed by Transit Authority for technical accuracy and their recommended corrections have been incorporated in this final draft. However, all conclusions presented in the report are solely the responsibility of the New York City Department of Transportation. They do not necessarily reflect the views of the Transit Authority or the Metropolitan Transportation Authority. For the purposes of the report, the Central Queens Corridor (through which subway lines pass to reach the Queens cordon) is defined as the geographic area running from Flushing Meadows Park in the east to Long Island City in the west, and from Rikers Island Channel in the north to the Long Island Railroad main line in the south. The map on the next page shows the corridor and its three subway lines. It also shows the two LIRR lines that pass through the corridor but do not attempt to serve local residents. 1 Click HERE for graphic. All data used in this report to describe "current conditions" reflect the Transit Authority's passenger and train counts of October 1989. This appeared to be the most suitable baseline for planning purposes. Passenger loadings are slightly lower than the peaks reached during the economic boom of the 1980's, but they do not reflect the temporary declines caused by the current recession. REPORT HIGHLIGHTS During the morning peak hour, the average subway train from Queens enters Manhattan overloaded by 10 percent. one-third of these trains are overloaded by more than 20 percent. Subway overcrowding causes a stressful commute for the CBD workers who must use these trains. This can reduce their productivity on the job and cost their employers money. Lack of peak hour reserve capacity on these subway lines limits the ability of Queens to provide more workers for CBD firms. Since Queens is one of the region's most important sources of ambitious and well-trained workers, this can inhibit long term job growth in the CBD. The report describes six options for providing enough new subway capacity to eliminate overcrowding and assure a comfortable reserve to accommodate future growth. Each option can stand alone. But they can also be combined in various scenarios that would provide incremental increases in capacity over many years to match long term growth in demand. Five of the options involve using the 63rd Street line to enable more peak hour trains to enter Manhattan from Queens. One of these is the MTA's proposal to connect the Queens Boulevard line to the 63rd Street line. The sixth option provides new capacity solely by running longer trains on existing subway lines. This would have the least impact on subway operating costs. 2 Two of the options using the 63rd Street line would extend subway service from midtown Manhattan into areas of Queens that are currently underserved or not served at all. They could be combined together to substantially increase subway coverage of Queens as well as providing more peak hour capacity. One of the options (the MTA's proposal) would also improve subway system reliability by providing an alternative route into Manhattan for Queens Boulevard line trains. This is an important benefit that the other options are not able to offer. CURRENT CONDITIONS 1. The Subway Lines Three subway lines run through various portions of the Central Queens Corridor and enter Manhattan across the Queens cordon. They are the Flushing IRT line, the Queens Boulevard IND line, and the Astoria BMT line. These lines use three two-track tunnels under the East River to access Manhattan. The 60th Street tunnel serves N trains from the Astoria line and R (local) trains from the Queens Boulevard line. The 53rd Street tunnel serves E and F (express) trains from the Queens Boulevard line. The 42nd Street tunnel serves No. 7 trains from the Flushing line. The new 63rd Street line provides a fourth tunnel under the East River. It connects with the Sixth Avenue IND line and Broadway BMT line in Manhattan and currently terminates at 21st Street in Long Island City. The MTA proposes to connect the 63rd Street line to the Queens Boulevard line as a first step in providing more subway capacity across the Queens cordon. 3 2. Subway Passenger Capacity The functional passenger capacity of these subway lines is defined by the following factors: The optimal passenger capacity of each subway car, as established by the MTA. IRT cars have a capacity of 110 passengers. IND cars that are sixty feet long (Classes R42 and earlier) have a capacity of 145 passengers. IND cars that are seventy-five feet long (Classes R44 and later) have a capacity of 175 passengers. For the purposes of this report, all Queens Boulevard line and Astoria line trains are assumed to be composed of seventy five foot cars. The number of cars per train. Constraints on maximum train length are imposed by the length of station platforms and the configuration of the signal system. Flushing IRT line trains can operate eleven car trains. The Queens Boulevard IND line and Astoria BMT line can operate eight car trains. The number of trains per hour that can be accommodated by each subway track. Textbook theory suggests that this is 30 trains per hour. But the practical realities of subway operations in New York City limit track capacity to 27 trains per hour. Given these factors, the 77 trains operating in-bound to Manhattan on all three lines during the morning peak hour are defined for the purposes of this report as having a functional capacity of 102,670 passengers. (The Transit Authority has been experimenting with alternatives to the October 1989 configurations of car types and mix of trains per hour on E and F trains in order to fine- tune capacity and match it more closely to demand. The results of the experiment, which is still in progress, are being evaluated by the TA. They are not reflected in this report, which is based on the October 1989 configurations.) 4 3. Overall Peak Hour Passenger Demand Under current conditions, all three subway lines together face total demand of 112,790 passengers during the morning peak hour. This means that their average "load factor" is 110 percent. (A subway line's load factor is a measure of how crowded its trains are. A load factor of 110 percent means that it is carrying ten percent more passengers than its functional capacity, and is therefore defined as it overcrowded". A load factor of 90 percent means that it is carrying ten percent fewer passengers than its functional capacity, and therefore has some reserve capacity.) Of the 77 Manhattan-bound peak hour trains on all three lines: * 84 percent have load factors higher than 100 percent (and are therefore defined as overcrowded) . * 49 percent have load factors higher than 110 percent (or more than the average load factor for all 77 trains). * 35 percent have load factors higher than 120 percent. * 17 percent have load factors higher than 130 percent. By comparison, only 16 percent of the 77 trains have load factors that are lower than 100 percent. These conditions have two results. One is that overcrowding can be severe on many trains. The second is that there is, on average, no reserve capacity to accommodate future growth in peak hour passenger flow across the Queens cordon. 5 4. Line-By-Line Peak Hour Passenger Demand Load factors for the individual train services that run on the three subway lines differ significantly from the average load factor given above. The bar chart on the next page shows the peak hour load factor for each train service, as well as the average load factor for all 77 trains crossing the Queens cordon. E, F, and N trains have load factors that are higher than the average. Taken together, they provide 52 percent of the functional capacity at the Queens cordon. But they carry 58 percent of the peak hour passengers into Manhattan. * The 14 peak hour E (Queens Boulevard line) express trains have load factors of 122 percent as they pass through the 53rd tunnel into Manhattan. They provide 19 percent of the functional capacity at the cordon, but carry 21 percent of peak hour passengers. These trains proceed west across Manhattan under 53rd Street and turn south onto the Eighth Avenue IND line. * The 13 peak hour F (Queens Boulevard line) express trains have load factors of 133 percent as they pass through the 53rd Street tunnel. They provide 18 percent of the functional capacity at the cordon, but carry 21 percent of peak hour passengers. These trains proceed west across Manhattan under 53rd Street and turn south onto the Sixth Avenue IND line. * The 11 peak hour N (Astoria line) trains have load factors of 112 percent as they pass through the 60th Street tunnel into Manhattan. They provide 15 percent of the functional capacity at the cordon and carry 15 percent of peak hour passengers. These trains proceed west across Manhattan under 59th Street and turn south onto the Broadway BMT line. Number 7 and R trains have load factors that are below the cordon's average. Together, they provide 48 percent of the functional capacity at the cordon and carry 42 percent of peak hour passengers. 6 Click HERE for graphic. * The 27 peak hour Number 7 (Flushing line) trains have load factors of 108 percent as they pass through the 42nd Street tunnel into Manhattan. They provide 32 percent of the, cordon's functional capacity and carry 31 percent of its peak hour passengers. These trains proceed west across Manhattan under 42nd Street and terminate at Eighth Avenue. * The 12 peak hour R (Queens Boulevard line) local trains have load factors of 72 percent as they pass through the 60th Street tunnel, which they share with N trains from the Astoria line. This is the only one of the five train services that has a load factor of less than 100 percent. These trains provide 16 percent of the cordon's functional capacity, but carry only 11 percent of its peak hour passengers. Like N trains, R trains proceed west across Manhattan under 59th Street and turn south onto the Broadway BMT line. The diagram on the next page shows the cordon-crossing routes for all five train services, along with their load factors. 5. The Load Balancing Problem E and F trains experience the most severe overcrowding problems. In part, this arises from an unwillingness on the part of Queens Boulevard line passengers to balance peak hour loads between E and F express trains and R local trains (which, as noted above, travel into Manhattan with load factors of only 72 percent). If the 60,156 peak hour passengers on the Queens Boulevard line were to allocate themselves evenly among the E, F, and R, each train would have a load factor of 110 percent. This would produce a significant (though not complete) reduction in overcrowding on E and F trains. If passengers were to allocate themselves so that R trains had load factors of no more than 100 percent (i.e. loads equal to capacity), E and F trains would have load factors of 114 percent. This is still an improvement over current conditions. It is not clear why passengers do not allocate themselves among these services more evenly. R trains make the same number of station stops between Queens Plaza and 34th Street as F trains, and one fewer stop than E trains. The route they follow in Manhattan is six blocks north of 7 Click HERE for graphic. the E and F during its westbound segment. After turning south, R trains run down Broadway between the E (on Eighth Avenue) and the F (on Sixth Avenue). Two theories offered to explain this uneven allocation are: Passenger resistance to changing their long established travel patterns. The perception that R trains are "slower" than E and F trains. Load balancing is an important issue that needs to be studied more closely. Most of the options described below for increasing subway capacity across the Queens cordon would require passengers to change their current travel patterns if the additional capacity is to eliminate overcrowding on Queens Boulevard line express trains. 6. Consequences of Subway Overcrowding * Overcrowding can hamper subway operations in ways that reduce functional capacity. - Overcrowded trains take longer to load and unload passengers when they are stopped at stations. The resulting increase in station "dwell time" can reduce the number of trains actually crossing the cordon during a given functional capacity. Since passenger demand does not change, fewer trains can lead to even worse overcrowding and still longer dwell times. - In general, overcrowded trains tend to have higher incidences of stuck doors and sick passengers. These random events can disrupt subway service for a quarter of an hour or more, which has a devastating impact on the smooth flow of passengers across the Queens cordon. * Since most peak period subway passengers are on their way to jobs in the Manhattan CBD, subway overcrowding causes a stressful commute that can 8 reduce the productivity of these workers. This has cost implications for firms located in the CBD. * Subway overcrowding during commuting periods can reduce the CBD's attractiveness as a place to work. This increases the difficulty of assuring an adequately large and experienced labor supply for the CBD, which has other cost implications for CBD firms. These firms may have to pay higher salaries than would otherwise be the case in order to obtain the number of qualified workers they require. * An average load factor in excess of 100 percent means that there is no reserve capacity on the subway lines affected to accommodate future growth. This can act as a constraint on the CBD's ability to accommodate more jobs in the future, since jobs cannot locate in the CBD if workers cannot get to them. Constraints on CBD job growth have important implications for New York City's future rate of economic growth. CBD jobs tend to have "higher value" than jobs located elsewhere in the five boroughs - in terms of the personal income and tax revenues they produce, plus the number of ancillary jobs they generate for support services (in restaurants, retail stores, etc.) both within the CBD and within the residential neighborhoods where CBD job-holders live. * Lack of reserve capacity limits opportunities to shift CBD auto commuters out of cars and into public transportation. Future shifts of this kind may be necessary in order to reduce motor vehicle traffic congestion and air pollution in Queens. It seems evident that subway overcrowding at the Queens cordon is a serious problem that needs to be addressed. One way to do this is to increase subway capacity. WHY INCREASE SUBWAY CAPACITY? Measures to materially increase subway capacity will have high capital costs, and most will add to operating costs. Therefore, it is necessary to assure that capacity increases generate sufficient benefits to justify these new costs. 9 As noted above, improving the quality of the daily commute for existing riders can have positive economic benefits for CBD workers and employers. Providing reserve capacity to accommodate increased passenger demand could produce even greater benefits. There are three reasons why passenger demand on Queens subways could increase in the future. * Projections by the New York Metropolitan Transportation Council (NYNTC) and others show significant increases in motor vehicle trip demand on Queens highways during the next twenty five years. But these highways are already overcrowded and cannot be expanded. In addition, federal mandates to improve air quality may require reductions in the current volume of motor vehicle trips during peak periods. increasing the proportion of trips between Manhattan and Queens that are made by public transportation rather than motor vehicles would address both issues. Most of these new public transportation trips would have to be accommodated by Queens subway lines. * The long run trend of economic activity in Manhattan could accelerate if New York City's economy becomes more closely linked to high-growth regions in the global marketplace. This would increase CBD job growth. Since a significant proportion of the region's labor force lives in Queens and the Long Island counties, these areas would be an important source of new workers to fill these jobs. Their Manhattan work trips would increase passenger demand on Queens subways during commuting periods. * As defense-oriented business activity on Long Island shrinks, a larger proportion of the work force residing in Queens, Nassau, and Suffolk may seek jobs in Manhattan. This would add further to commuter demand on Queens subways. If the subway system is unable to accommodate this new demand, the economic activity that generates it will locate elsewhere and produce no benefits for New York City. Given current subway over- crowding at the Queens cordon, more subway capacity seems essential if new demand is to be accommodated. 10 OPTIONS FOR INCREASING SUBWAY CAPACITY There are three ways to increase subway capacity at the Queens cordon. * Run longer (therefore higher capacity) trains on the train services that are overcrowded. This has two advantages. One is that the new capacity is closely targeted to the train services that are overcrowded, which means that passengers do not have to change their existing travel patterns to make use of it. The second advantage is that longer trains would have only a modest impact on subway operating costs, since there is no need to increase the number of train crews or other operating personnel. * Build the kind of new connections between existing subway lines that enable more trains to cross the Queens cordon. This was done in the early 1950's when the local tracks of the Queens Boulevard line (which originally had no access to Manhattan) were connected to the 60th Street tunnel. During the 1990's, the MTA proposes to connect both the local and express tracks of the Queens Boulevard line to the 63rd Street line, which provides a new river tunnel into Manhattan. * Build new subway lines in Queens with new access tunnels to Manhattan. In this case, the 63rd Street line already provides the new river tunnel and suitable connections to existing subway lines in Manhattan. All that remains is to build the new subway lines in Queens. The six options described below provide at least one example of each approach for increasing peak hour subway capacity at the Queens cordon. The table on the next page compares their key features. The report does not evaluate options involving augmented commuter rail service or new systems using trolley or people mover technologies, since they are outside its scope. But their exclusion does not necessarily imply that they lack merit. Implementing any one of these subway options does not preclude implementing any (or several) of the others at a some later time. Various scenarios are possible for phasing in several options in sequence over a period of many years. Doing so would provide incremental increases in capacity to match incremental growth in demand. 11 Click HERE for graphic. An important factor in evaluating these options is the proposal to build a new passenger station in Sunnyside Yard for Amtrak and LIRR trains. Commuter access to Long Island City would be further enhanced by convenient passenger transfers between this station and the subway lines passing nearby. This would help stimulate its development into a major commercial center. All else being equal, subway options that can provide transfers to this station are highly desirable. The report does not consider whether the options could also be used to improve landside access for air passengers using JFK and LaGuardia airports. Air passengers have special needs and characteristics. It has become increasingly apparent that it is neither feasible nor desirable to try and serve them by subway. Other transit options for improving landside access to the airports are being explored by the Port Authority and the City. THE SIX OPTIONS, 1. Run Longer Trains Running longer trains appears to be a feasible (though not necessarily simple) option for each of the overcrowded train services. No changes in existing passenger travel patterns would be necessary for the new capacity to eliminate overcrowding. But passengers would have to distribute themselves evenly among the cars of each train to avoid overcrowding in certain cars. The diagram on the next page shows the impact of this option. On the Flushing line, extending train lengths from eleven cars to twelve cars would reduce peak hour load factors to 99 percent under current demand conditions. Extending train lengths to fourteen cars would reduce load factors to 85 percent. This would leave a capacity reserve of 6,160 passengers per hour to accommodate future growth. On Queens Boulevard line E expresses, extending train lengths from eight cars to ten cars would reduce peak hour load factors to 98 percent under current demand conditions. Extending train lengths to eleven cars would reduce load factors to 89 percent. This would leave enough reserve capacity to accommodate demand growth of 2,983 new passengers per hour. 12 Click HERE for graphic. * On Queens Boulevard F expresses, extending trains lengths from eight cars to eleven cars would reduce peak hour load factors to 96 percent under current demand conditions. Extending train lengths to twelve cars would reduce load factors to 88 percent. The reserve capacity available would be able to accommodate demand growth of 3,155 additional passengers per hour. * On the Astoria line, extending the length of N trains from eight to nine cars would reduce peak hour load factors to 99 percent under current demand conditions. Extending train lengths to ten cars would reduce load factors to 89 percent. This would allow for demand growth equal to 2,036 passengers per hour. To accommodate longer trains, station platforms would have to be extended and signal systems modified on the subway lines affected. * For Flushing line trains, these modifications would be limited to the Flushing line itself. * For E trains, the modifications would also have to include the Eighth Avenue line in Manhattan between 50th Street and the World Trade Center. * For F trains, the modifications would also have to include the Sixth Avenue line in Manhattan and the Culver line in Brooklyn. * For N trains, the modifications would also have to include the Broadway line in Manhattan, and the Fourth Avenue and Sea Beach lines in Brooklyn. More subway cars would have to be purchased to fill out the longer trains. Storage tracks and maintenance facilities serving these car fleets would also have to be modified. By providing more capacity on train services that are now overcrowded, this option would enhance any transfer provisions made between existing subway lines and the proposed Amtrak/LIRR station in Sunnyside Yard. 13 2. Connect The Oueens Boulevard Line To The Upper Level of The 63rd Street Line The MTA has proposed to build a connection between these two lines at a point east of Queens Plaza station. This connection would provide two important benefits. * It would give Queens Boulevard line trains a second route to and from Manhattan when their main route is temporarily blocked. In statistical terms, subway service is highly reliable. But random blockages due to stuck train doors, sick passengers, etc. are not uncommon because of the large number of trains in operation. The existence of an alternate route can prevent such blockages from causing a total breakdown in peak hour service. * It would make possible a 36 percent (14 trains per hour) increase in the number of Queens Boulevard line trains to and from Manhattan. The total number of morning peak hour trains crossing the Queens cordon (on all lines) would increase by 18 percent, from 77 to 91. This 19,600 passenger per hour increase in functional capacity would reduce the average load factor at the cordon to 92 percent under current demand conditions, leaving a 9,480 passenger per hour reserve to accommodate future growth. The MTA has been running commuter simulations of various train service patterns that would be possible on the Queens Boulevard line to make use of the additional track capacity into Manhattan. The diagram on the next page shows the projected impact of one of these service patterns. The MTA's simulations indicate that passengers would have to change their travel patterns, possibly in significant ways, in order to balance loads so that the additional capacity would eliminate overcrowding. Otherwise, some express trains will continue to have load factors greater than 100 percent while load factors on local trains remain below 100 percent. The 14 additional Queens Boulevard line trains would use about half of the track capacity on the 63rd Street line's upper level. The rest would remain available for use in the future by other options. 14 Unlike longer trains, this option would increase subway capacity by running more trains. Doing so requires more train crews and would mean higher subway operating costs. Because of the location of the track connection, trains operating via 63rd Street would not be able to offer passenger transfers to the proposed Amtrak/LIRR station in Sunnyside Yard. 3. Reverse Signaling For The Oueens Boulevard Line This option assumes that the connection described above is built. It provides a way to utilize the full track capacity of the 63rd Street line's upper level by Queens Boulevard line trains exclusively. The Queens Boulevard line's two express tracks would be resignaled for bi-directional operation. This would allow both express tracks and one local track to run Manhattan-bound trains during the morning commuting periods (a "3 & 1" service pattern), and just the reverse in the afternoon. By effectively adding one additional express track to the Queens Boulevard line in the direction of primary demand, this option would double the number of Manhattan-bound express trains during the morning peak hour - from 27 to 54. The line's passenger capacity (in the direction of primary demand) would be increased by 69 percent, or 37,800 passengers per hour. The diagram on the next page shows the impact of one service pattern that is possible with this option. All of the new capacity would be in express trains. Half of them would enter Manhattan via the 53rd Street tunnel. The other half would enter via the 63rd Street tunnel ' Under current demand conditions and with no change in the present allocation of passengers between expresses and locals, these express trains would have an average peak hour load factor of 64 percent. But it is possible that load factors could differ considerably between express trains taking different routes into Manhattan unless passengers engage in aggressive load balancing. With this option, the total functional capacity of all train services crossing the Queens cordon would be 37 percent greater than the current level. The average peak hour load factor for all trains would be 80 percent. This would provide a capacity reserve of 27,680 passengers per hour to accommodate future growth. 15 Click HERE for graphic. The success of this option depends heavily on overcoming the practical complications of operating a 3 & 1 service pattern during peak periods. These complications may be substantial and should be evaluated with care. Since all of the new capacity would be provided by running more trains, operating costs would be higher than now. As noted above, none of the express trains operating via the 63rd Street line would be able to offer passenger transfers to the proposed Amtrak/LIRR station in Sunnyside Yard. 4. Connect The Flushing Line To the 63rd Line This connection would mean converting the Flushing line (which now operates IRT trains) to IND standards so that its trains could operate on the Sixth Avenue IND and Broadway BMT lines in Manhattan. The diagram on the next page shows its potential impact. Because IND trains are longer and wider than IRT trains, such a conversion would increase the Flushing line's functional capacity by 16 percent, or 5,130 passengers per hour. This would reduce the line's peak hour load factor to 94 percent and eliminate its present overcrowding problems. It could also absorb the excess passengers on N trains (by means of transfers between the two train services at the Queensboro Plaza station). Therefore, peak hour overcrowding on both N and Flushing line trains could be eliminated. This option would increase total functional capacity at the Queens cordon by five percent. Therefore the average peak hour load factor for all trains would be reduced to 105 percent from the present 110 percent. This option would utilize the entire track capacity in the 63rd Street line's upper level and would have only minor operating cost implications (new capacity would be provided by running higher capacity trains rather than more trains). It would also enhance the value of any transfer provisions made between the Queensboro Plaza station and the proposed Amtrak/LIRR station in Sunnyside Yard. The Flushing line could even be provided with its own transfer station beneath the Amtrak/LIRR station, as part of its connection to the 63rd Street line. 16 Click HERE for graphic. But this option would not be able to provide enough new capacity by itself to eliminate all overcrowding at the Queens cordon and to provide any reserve to accommodate future growth. To accomplish both goals, it would have to be combined with longer express trains on the Queens Boulevard line. Connecting the Flushing line to the 63rd Street line would free up the 42nd Street tunnel under the East River and across Manhattan for other uses. These might include serving as the right-of-way for: * A new trolley ("light rail") line between midtown Manhattan and Long Island City. * A new subway line between New Jersey and Long Island City via midtown Manhattan, which is being evaluated by the MTA and the Port Authority. * An airport transit system using people mover technology to connect JFK and LaGuardia with midtown Manhattan. 5. Connect The- LIRR's Port Washington Line To the Upper Level Of The 63rd Street Line If the Port Washington line was converted to IND standards, it could be connected to the 63rd Street line and become part of the subway system. This option would have the following characteristics. * An entirely new subway line would be provided through the Central Queens Corridor to Flushing, on through northeast Queens, and into Nassau. This would extend subway service to areas that have never been served before. * Enough subway trains could be operated on this line to utilize the full track capacity of the 63rd Street line's upper level. * Alternatively, this option could be combined with the Queens Boulevard line connection to 63rd Street. The two lines would then share the track 17 capacity of 63rd Street's upper level. * In building the connection to 63rd Street, it may be possible to provide the line with its own transfer station beneath the proposed Amtrak/LIRR station in Sunnyside Yard. * The LIRR's seven peak hour Port Washington trains would be removed from Penn Station. This would provide new platform capacity for more LIRR trains from other lines. * The line could (and might have to) be operated as a premium fare "super subway" with more accommodation for seated passengers. * After allowing for more seated passengers per car and the need to accommodate the Port Washington line's existing LIRR passengers, the line could increase subway capacity at the Queens cordon by 19 percent. Under current demand conditions, this would reduce the average peak hour load factor for all trains at the cordon to 93 percent. A capacity reserve of 9,450 passengers per hour would be available to accommodate future growth. Operating the line as a premium fare super subway may be both necessary and beneficial. Necessary, because of the social and political implications of extending the subway system into up-scale areas of northeast Queens and northwest Nassau. Beneficial, because of its potential for attracting passengers to the subway system - as opposed to simply reallocating existing passengers. The line's higher amenity level (vs. other subway lines) plus more frequent service and better distribution capability within Manhattan (vs. the LIRR) would be the key elements in this potential. As the diagram on the next page illustrates, the line's trains could operate along both the Sixth Avenue IND and Broadway BMT lines in Manhattan. They would be prominently color-coded to distinguish them from regular subway trains. Implementation of automated fare card technology in the subway system would greatly simplify the collection of premium fares. Passengers outbound from Manhattan would pay the regular fare to enter stations on Sixth Avenue or Broadway, and pay the rest of the premium fare when they exit stations in northeast Queens or Nassau. 18 Click HERE for graphic. Available evidence suggests that significant numbers of Flushing line and Queens Boulevard line passengers now use local bus lines to access them from areas (like northeast Queens) that the subway system does not currently serve. Modifications to these bus routes might enable them to feed passengers to the Port Washington line. This could be an effective way of reallocating passenger loads so that the new capacity provided by the Port Washington line can eliminate overcrowding on the Flushing and Queens Boulevard lines. The map on the next page shows the route of this line in Queens. 6. Build A New Oueens Subway The 63rd Street line was originally conceived as the Manhattan entry for an entirely new subway line in Queens. This line would have extended from the end of the 63rd Street line in Long Island City, through Sunnyside Yard and along the LIRR main line right-of-way to Rego Park, where it would have entered a short new subway tunnel to connect with the Queens Boulevard line at Continental Avenue. It is possible to revive this concept in a way that both avoids the high cost of building underground track connections at Continental Avenue and enables the line to serve a much larger portion of Queens. From the present end of the 63rd Street line in Long Island City, this new subway line would run through Sunnyside Yard and along the LIRR main line right- of-way to Rego Park. There, it would turn south onto the abandoned right-of-way of the LIRR's old Rockaway branch and run to a connection with the Rockaway IND subway line at Liberty Avenue in Ozone Park. This option would have the following characteristics. * An entirely new subway line would be provided through the Central Queens Corridor to Rego Park, south through Woodhaven and Ozone Park, and on to the Rockaways. * Enough subway trains could be operated on this line to utilize the full track capacity of the 63rd Street line's upper level. 19 Click HERE for graphic. * Alternatively, this option could be combined with the Port Washington option (or the Queens Boulevard line connection) to share the track capacity of 63rd Street's upper level. The combination with the Port Washington option would provide Queens with substantially more (and more widely distributed) subway service. * In building the connection to 63rd Street, it may be possible to provide the line with its own transfer station beneath the proposed Amtrak/LIRR station in Sunnyside Yard. * By providing faster and more direct service between the Rockaways and midtown Manhattan, the line might stimulate major new development in Arverne and other underdeveloped areas in the eastern portion .of the Rockaway peninsula. * If the line was the sole connection to 63rd Street, its 27 trains per hour would increase subway capacity at the Queens cordon by 37 percent. Under current demand conditions, this would reduce the average peak hour load factor at the cordon to 80 percent. A capacity reserve of 27,680 passengers per hour would be available to accommodate future growth. The map on the next page shows the route of this line in Queens, along with its connection to the Rockaway line plus a connection to the Jamaica Avenue line that would also enable it to serve the Jamaica business center. It is followed by a diagram of one possible service pattern. COMPARISONS Six factors can be used to compare each of the options on a stand-alone basis. They are: (1) the amount of new capacity provided; (2) the need for changes in passenger travel patterns in order to fully utilize this capacity; (3) capital and operating costs; (4) improvements to subway system reliability; (5) the ability to provide service from midtown Manhattan to underserved areas in Queens; and (6) the opportunities for passenger transfers to the proposed Amtrak/LIRR station in Sunnyside Yard. 20 Click HERE for graphic. Click HERE for graphic. New Capacity * The new Queens subway and reverse signaling on the Queens Boulevard line (with the connection to 63rd Street) both provide the most new capacity. Each would increase in functional capacity at the Queens cordon by 37 percent. * Running longer trains would increase capacity by 29 percent. * The Port Washington connection would increase capacity by 28 percent. * The Queens Boulevard connection (without reverse signaling) would increase capacity by 18 percent. * The Flushing line connection would increase capacity by five percent. 2. Travel Pattern Changes Needed * Running longer trains and the Flushing line connection would require no changes in passenger travel patterns for the new capacity to eliminate overcrowding on the train services targeted. * Reverse signaling on the Queens Boulevard line would probably require no changes in passenger travel patterns to eliminate overcrowding on express trains. But significant changes would be needed to eliminate overcrowding on Flushing line trains, since passengers would have to shift to Queens Boulevard line express trains in order to balance loads between the two lines. * The Queens Boulevard connection (without reverse signaling), the Port Washington connection, and the new Queens subway would each require significant changes in travel patterns to eliminate overcrowding. 21 3. Capital And Operating Costs (The Queens Boulevard connection is the only option for which detailed capital cost estimates currently exist.) * Running longer trains might have the lowest capital cost if the resignaling on the lines affected is done as part of normal replacement of existing signal systems that have reached the end of their useful life. This option would have the lowest operating cost impact because it does not involve running more trains. * The capital cost of the Queens Boulevard connection is estimated by the MTA at about $670 million, which is likely to be near the low end of the scale for the six options. But it would have a material operating cost impact because the new capacity is provided by running 14 more peak hour trains. * The Flushing line connection would have a higher capital cost than the Queens Boulevard connection because of the need to convert the line to IND standards. But its operating cost impact would be modest because its new capacity does not involve running more trains. * Reverse signaling on the Queens Boulevard line (including the connection to 63rd Street) would probably have the next highest capital cost unless resignaling can be done as part of normal signal replacement. Its operating cost impact would be higher than for the Queens Boulevard connection alone because its new capacity is provided by running 27 more trains per hour rather than 14 more trains. The complications of running a "3 & 1" service pattern during commuting periods could also add to the operating cost impact. * The Port Washington connection would have the next highest capital cost because of the need to convert an LIRR line to IND standards. Its operating cost impact would be significant because it adds what amounts to a lengthy new subway line to the system. 22 * The new Queens subway would have the highest capital cost because it involves building an entirely new subway line (mostly on the surface) from the end of the 63rd Street line in Long Island City to the connection with the Rockaway line in Ozone Park. its operating cost impact would probably be about as high as the Port Washington connection. 4. Improving Subway System Reliability * The Queens Boulevard connection (with or without reverse signaling) is the only option that would have a significant impact on improving the operating reliability of the subway system. It would provide an additional route to Manhattan for Queens Boulevard line trains, which offers more operating flexibility during temporary line blockages. * Reverse signaling, the Flushing line connection, the Port Washington connection, and the new Queens subway might reduce reliability slightly because they would add more trains to subway lines in Manhattan. More trains can mean increased incidences of stuck doors, sick passengers, and similar incidents that are inevitable on such a complex and heavily used subway system. But any changes in reliability would probably be so minor that passengers would be unaware of them. 5. Subway Service To New Areas From Midtown Manhattan * The Port Washington connection would provide subway service for the first time to northeast Queens and northwest Nassau. * The new Queens subway would provide more direct service from midtown Manhattan to the Rockaways. Present service (provided by A and C trains) runs from midtown through lower Manhattan, then through downtown Brooklyn, then through central and eastern Brooklyn, then on to the Rockaways. 23 * None of the other options extend subway service from midtown Manhattan to areas not presently served. (The map on the next page shows the impact of connecting both the Port Washington line and the new Queens subway to the 63rd Street line. Queens would be provided with a substantial expansion of subway service, flowing eastward out of 63rd Street and dividing to reach both northeast Queens and the Rockaways.) 6. Transfers To the Proposed Amtrak/LIRR Station * The Flushing line connection, the Port Washington connection, and the new Queens subway could probably be provided with a direct transfer to this station as part of their connections to the 63rd Street line. * Running longer trains would enhance any transfer provisions made between existing subway lines and this station. * The Queens Boulevard connection (with or without reverse signaling) would provide no transfer opportunities to the proposed station for passengers on trains using the 63rd Street line. 24 Click HERE for graphic. II. C U R R E N T C 0 N D I T I 0 N S THE CENTRAL QUEENS CORRIDOR For the purposes of this report, the Central Queens Subway Corridor is defined as the funnel-shaped geographic area lying between the Long Island City shore of the East River (the narrow end of the funnel) and Flushing Meadows Park (the tunnel's wide end). It includes the communities of Long Island City, Sunnyside, Woodside, Jackson Heights, Elmhurst, Rego Park, Corona, and Forest Hills. As the map on the next page indicates, two major subway lines and two LIRR lines run roughly east to west through the corridor. A third subway line (the Astoria line) enters the western end of the corridor from the north, but doesn't directly serve most of the corridor. The three subway lines are: 1. The Flushing IRT Line This line enters the northern portion of the corridor from the east near Shea Stadium. It proceeds generally west on an elevated structure above Roosevelt Avenue to Queens Boulevard, then generally west on an elevated structure above Queens Boulevard to and across Sunnyside Yard to Queens Plaza. Then it proceeds generally south to Hunters Point Avenue, where it turns west again and enters the tunnels that carry it under the East River to 42nd Street in Manhattan and west to a terminal between Seventh and Eighth avenues. The line has two main tracks throughout its length. But between Shea Stadium and Woodside,, it also has a third (center) track. This is used to provide express service in the direction of primary demand during commuting periods. 2. The Oueens Boulevard IND Line This four track subway line enters the southern portion of the corridor from the east at the southeast corner of Forest Hills. It proceeds generally west in a subway under Queens Boulevard to Elmhurst, then generally west in a subway under Broadway to Woodside, then generally west in a subway under Northern Boulevard to Queens Plaza in Long Island City. 25 Click HERE for graphic. From Queens Plaza, its express tracks proceed generally west to the tunnels that carry them under the East River to 53rd Street in Manhattan. These tracks continue west under 53rd Street to Sixth Avenue (and a connection with the Sixth Avenue IND line) and to Eighth Avenue (and a connection with the Eighth Avenue IND line). At Queens Plaza, the line's local tracks divide into two routes. One connects with the Astoria BMT line to enter Manhattan through the 60th Street tunnel under the East River. This route continues west under 60th and 59th streets to Seventh Avenue, where it turns south and connects with the local tracks of the Broadway BMT line. The second route turns south from Queens Plaza and runs through Greenpoint, Williamsburg, and Bedford-Stuyvesant to a connection with the Fulton Street IND line in downtown Brooklyn. This route does not enter Manhattan. As noted, this line has four tracks as far west as Queens Plaza and runs underground throughout. From an operating engineering standpoint, it is built to very high standards - probably the highest of any subway line in New York and among the highest in the world. Its station platforms are long and wide, there are a number of storage sidings along the way to accommodate temporarily disabled trains, and none of its tracks cross each other at grade. It was built in the 1930's and was New York's last major trunk line subway running into Manhattan to be constructed from scratch. 3. The Astoria BMT Line This line enters the corridor from the north at its extreme western end. It runs on an elevated structure from Astoria to Queens Plaza, where it shares a cross-platform passenger transfer with the Flushing line at the Queensboro Plaza elevated station. west of Queens Plaza, the line enters the 60th Street river tunnel (which it shares with local trains from the Queens Boulevard line) to reach Manhattan. The two LIRR lines also run east to west through the corridor but do not attempt to serve corridor residents. These lines are: 4. The LIRR Main Line This line enters the southern portion of the corridor from the east at a point several blocks south of the Queens Boulevard subway line. It proceeds generally west 26 on an elevated embankment that roughly parallels the Queens Boulevard line to Sunnyside Yard and the tunnels that carry it under the East River to Penn Station in Manhattan. In Woodside and Hunters Point, it provides transfers to the Flushing line for LIRR passengers desiring a more direct route to the eastern portions of Midtown Manhattan. 5. The LIRR Port Washington Line This line enters the northern portion of the corridor from the east near Shea Stadium, several block south of the Flushing line. It proceeds generally west on an elevated embankment that roughly parallels the Flushing line to Woodside, where it joins the LIRR main line and follows it into Penn Station. Like the main line, it provides a passenger transfer at Woodside. TECHNICAL CONSIDERATIONS Passenger crowding on subway lines at the Queens cordon is a function of both travel demand and the capacity of these lines. * Travel demand is a function of land use, demographic, employment, and residential patterns within the corridor. A detailed discussion of the factors affecting demand is outside the scope of this report. * Capacity is a function of technical factors that are built into the subway lines. These factors are discussed below. The passenger capacity of a subway line is normally expressed in terms of the number of passengers it can carry per hour in one direction past some fixed reference point (such as a cordon crossing). This measure of capacity is determined by four factors. 1. Number Of Tracks In Each Direction Each of the four subway tunnels crossing the East River hag two tracks - one for each direction. However, the new 63rd Street tunnel also has a two track lower level that was designed to accommodate LIRR trains. Very expensive track connections on either side of the East River would have to be built before trains can use the lower level. 27 2. Number Of Trains Per Hour Per Track Textbooks traditionally assume that a single subway track can accommodate a maximum of 30 trains per hour. This provides an average headway between trains of two minutes. However, the Transit Authority has found from experience that the practical track capacity on most of its lines is 27 trains per hour, which lengthens the average headway between trains to two minutes and 13 seconds. This ten percent lower track capacity is due to various factors. One of the most important may be excessive crowding on trains during peak periods, which increases the time that trains must wait in stations ("dwell time") while passengers struggle to wedge themselves into overcrowded trains. All capacity calculations in this report assume the more conservative (and more realistic) track capacity of 27 trains per hour. 3. Number Of Cars Per Train The Flushing line operates 11 car trains (one car more than other IRT lines). Its trains are 550 feet long. The Queens Boulevard and Astoria lines can operate trains of either eight 75 foot long cars (Classes R44 and later) or ten 60 foot long cars (Classes R42 and earlier). Both types of trains are 600 feet long. Train lengths are limited by two factors. * The length of station platforms, which can sometimes be extended relatively inexpensively. * The length of signal blocks, which is very expensive to alter. The Transit Authority has recently adopted 67 feet as the standard car length for future car purchases. Nine car trains of these cars would be 603 feet long, which can be accommodated by existing station platforms and signal systems on IND and BMT lines. 28 4. Passenger Capacity Per Car The MTA has developed objective standards to measure the theoretical passenger capacity for its three types of subway cars. These standards assume that each seat is occupied and each standing passenger has not less than three square feet of floor space. Since subway passengers will tolerate less than this minimum three square feet of floor space, the practical capacity of subway cars is somewhat greater than their theoretical capacity. But when cars are loaded to more than their theoretical capacity, they are defined as being overcrowded. IRT cars, which are 50 feet long and nine feet wide, have a theoretical capacity of 110 passengers. IND cars of Classes R44 and later, which are 75 feet long and ten feet wide, have a theoretical capacity of 175 passengers. IND cars of Classes R42 and earlier, which are 60 feet long and ten feet wide, have a theoretical capacity of 145 passengers. Given these four factors, the Flushing IRT line has a functional capacity in each direction of 32,670 passengers per hour (110 passengers per car times 11 cars per train times 27 trains per hour). With eight car trains of 75 foot cars (Class R44 and later), the Astoria line and the local and express tracks of the Queens Boulevard line each have a functional capacity in each direction of 37,800 passengers per hour (175 passengers per car times eight cars per train times 27 trains per hour). With ten car trains of Class R42 and earlier 60 foot cars, each of these lines would have a functional capacity of 39,150 passengers per hour (145 passengers per car times ten cars per train times 27 trains per hour). Note that IND trains of the older cars have 3.6 percent more capacity than trains of the newer cars. Both types of trains are the same length (600 feet). But trains composed of the newer cars provide more seats, which take up more floor space and reduce the amount of space available for standing passengers. This accounts for their slightly lower passenger capacity. For the purposes of this report, all trains on the Queens Boulevard and Astoria lines are assumed to operate eight car trains of 75 foot cars. This was the case in October 1989, which was used as the baseline for measuring what the report defines as "current conditions". 29 After October 1989, the Transit Authority began experimenting with different configurations of cars per train and the mix of trains per service on E and F trains. Both train services were assigned ten car trains of sixty foot cars 'in order to provide slightly more passenger capacity. Since all IND cars have four doors per side (regardless of their length), this change also increases the number of doors per train by 25 percent - which facilitates more rapid passenger loading and unloading at stations during peak periods. This experiment is still underway, so the report is not able to reflect its results. THE MACRO PICTURE AT THE OUEENS CORDON Three tunnels currently provide subway service into Midtown Manhattan from the Central Queens Corridor. From north to south, they are: 1. The 60th Street tunnel, served by N trains from the Astoria line and R trains from the Queens Boulevard line. Currently, 23 Manhattan-bound trains (11 Ns and 12 Rs) pass through this tunnel during the 8 to 9 Am peak hour. This provides a functional capacity of 32,200 passengers, or 31 percent of the total capacity at the Queens cordon. 2. The 53rd Street tunnel, served by E and F trains from the Queens Boulevard line. Twenty-seven Manhattan-bound trains (14 Es and 13 Fs) operate during the Am peak hour, providing a functional capacity of 37,800 passengers. This is 37 percent of the total capacity at the Queens cordon. (Note: these baseline numbers do not reflect the Transit Authority's on-going experiment with slightly higher capacity trains and a scheduled mix of 12 E trains and 18 F trains.) 3. The 42nd Street tunnel, served by Number 7 trains from the Flushing line. Twenty-seven Manhattan-bound trains operate during the Am peak hour, providing a functional capacity of 32,670 passengers. This is 32 percent of the total capacity at the Queens cordon. Therefore, 77 trains enter Manhattan through all three tunnels during the morning peak hour. This provides a functional capacity of 102,670 passengers. Transit Authority passenger counts from October 1989 are used to measure actual passenger demand. Although the recession has temporarily reduced demand somewhat, October 1989 counts provide a 30 more realistic baseline for planning purposes. These counts show 112,790 passengers on these trains during the morning peak hour as they left the last station in Queens to enter the tunnels. This represents an average load factor for all 77 trains of 110 percent, or 10,120 passengers more than functional capacity. In gross and somewhat oversimplified terms I eight additional IND trains would be needed to eliminate this overcrowding. They would provide space for 11,200 more passengers per hour and would raise peak hour functional capacity at the Queens cordon to 113,870 passengers. This would reduce the average load factor for all trains to 99 percent. In order to eliminate overcrowding plus provide a ten percent reserve for future growth, 16 additional IND trains would be needed. They would provide space for 22,400 more passengers per hour, raising peak hour functional capacity at the Queens cordon to 125,070 passengers. This would allow 12,280 additional peak hour passengers to be accommodated without exceeding functional capacity. This macro picture is shown in the Total column of the table on the next page. THE MICRO PICTURE AT THE OUEENS CORDON The preceding analysis provides a rough sense of subway overcrowding at the Queens cordon and how many additional trains would be needed to eliminate this problem. But the analysis is over-simplified in two ways. * It implies that each tunnel has the same passenger capacity. We know this is not the case because the 42nd Street tunnel operates lower-capacity IRT trains. * It implies that each of the five train services operating through these tunnels is equally overcrowded. As the other columns of the table on the next page indicate, this is not the case and the variations are quite significant. Therefore, it is necessary to examine each of the lines separately. The table shows the details for each line and each train service (as of October 1989). The two diagrams following the table presents this information in graphic form. 31 Click HERE for graphic. Click HERE for graphic. Click HERE for graphic. The Flushing Line The Flushing line's 27 morning peak hour trains to Manhattan have a functional capacity of 32,670 passengers. The Transit Authority"s actual counts show a demand level of 35,420 passengers at the cordon crossing. This represents a load factor of 108 percent (2,750 excess passengers) and accounts for 31 percent of total peak hour Manhattan-bound demand at the Queens cordon. As noted earlier, the Flushing line shares a cross-platform passenger transfer with the Astoria line at the elevated Queensboro Plaza station. In effect, the Flushing line offers passengers willing to transfer at this point two different routes into Manhattan. The 11 morning peak hour N trains to Manhattan appear to enter the Queensboro Plaza station with a certain amount of reserve capacity. These trains have a functional capacity of 15,400 passengers. At the station, some Flushing line passengers transfer to N trains. As a result, N trains leave Queensboro Plaza (their last station in Queens) and cross the cordon during the peak hour with 17,214 passengers. This represents a load factor of 112 percent (1,814 excess passengers) and accounts for 15 percent of total demand at the Queens cordon. Therefore, these two routes into Manhattan, which have a combined peak hour capacity of 48,070 passengers, must accommodate combined demand levels of 52,634 passengers. Their load factors average 109 percent (4,564 excess passengers) and they account for 47 percent of total demand at the Queens cordon. It seems apparent that, by transferring at Queensboro Plaza, Flushing line passengers engage in a judicious amount of load balancing. Even so, both routes into Manhattan are overcrowded during the morning peak hour. This overcrowding might be alleviated by increasing the number of Manhattan-bound N trains. The 60th Street tunnel can accommodate 27 trains per hour in each direction, but is only operating 23 (11 N trains and 12 R trains) during the morning peak hour. Adding four N trains would raise this to 27. These four N trains would have a passenger capacity of 5,600 passengers, raising the combined capacity of Flushing line trains and N trains to 53,670 peak hour passengers. If passengers balanced loads by transferring at the Queensboro Plaza station, overcrowding would be eliminated. Flushing line 32 trains would have load factors of 98 percent at the cordon crossing and N trains would have load factors of 97 percent. Both lines together would an average load factor of 98 percent. However, there would be less than a two percent reserve to accommodate future growth. If a ten percent reserve was desired, the combined capacity of these two routes into Manhattan would have to be raised to 57,897 passengers. This could be done by adding three more N trains, which would require eliminating three R trains from the 60th Street tunnel. Such a reduction in R service would still leave R trains (which are currently under-utilized) with some reserve capacity at the cordon crossing. But their average headways would be increased by one-third, from five minutes to six minutes and 40 seconds. A more serious problem with this scenario is the assumption that Flushing line trains could accommodate the additional demand east of the Queensboro Plaza station. It is not clear that this could be done without subjecting passengers to severe overcrowding - which could lead to longer dwell times at stations and a reduction in the actual number of trains per hour reaching Queensboro Plaza. The functional effect of this would be less passenger capacity on the Flushing line. 2. The Oueens Boulevard Line Overcrowding on this line is much worse and more complicated than on the Flushing line. Part of the problem seems to be the reluctance of passengers to make effective use of R trains, which provide local service and have a different route into Manhattan than express trains. The Queens Boulevard line is a four track subway between Jamaica and Long Island City. In Jamaica, the line has a new two track branch that parallels the original four track line (in effect, six tracks from Jamaica merge into four tracks). West of Queens Plaza, the express tracks and local tracks follow different routes into Midtown Manhattan. For analytical purposes, the express tracks and local tracks can be considered as two separate lines that happen to share the same subway structure under Queens Boulevard and offer passengers cross-platform transfers at four express stations in the Central Queens Corridor. 33 * Express Service The express tracks have two train services. trains cross Manhattan under 53rd Street and turn south onto the Eighth Avenue IND line. trains also cross Manhattan under 53rd Street, but turn south onto the Sixth Avenue IND line. During the morning peak hour, the 14 E trains crossing the Queens cordon to Manhattan have a functional capacity of 19,600 passengers. Transit Authority counts show current demand of 23,967 passengers, or 21 percent of total peak hour Manhattan- bound demand at the Queens cordon. This represents a load factor of 122 percent (4,367 excess passengers). Four additional E trains would be needed to eliminate overcrowding. If these trains could be run (which track capacity in the 53rd Street tunnel does not permit), capacity would rise to 25,200 passengers and current demand would enable E trains to operate at load factors of 95 percent. The 13 F trains crossing the Queens cordon during the morning peak hour have a functional capacity of 18,200 passengers. Current demand is 24,145 passengers, or 21 percent of total demand at the Queens cordon. This represents a load factor of 133 percent (5,945 excess passengers). Five additional F trains would be needed to eliminate overcrowding. If these trains could be run (again, they cannot because of constraints in the 53rd Street tunnel), capacity would rise to 25,200 passengers and current demand would enable F trains to operate at load factors of 96 percent. Taken together, the two express services have a morning peak hour capacity into Manhattan of 37,800 passengers. Current demand is 48,112 passengers, or 43 percent of total demand at the Queens cordon. This represents an average load factor of 127 percent (10,312 excess passengers). If the nine additional E and F trains needed to eliminate overcrowding could be run (requiring the 53rd Street tunnel to accommodate 36 trains during the peak hour, which is not possible), capacity would rise to 50,400 passengers. Current demand would enable express trains to operate at average load factors of 95 percent. 34 If a ten percent capacity reserve was desired to accommodate future growth, 38 Manhattan-bound trains would be, needed during the morning peak hour (11 more than now). They would provide total capacity for 53,200 passengers. If demand grew by ten percent, E and F trains would have average load factors of 99 percent. * Local Service As noted earlier, the local tracks follow a different route into Manhattan via the 60th Street tunnel (which they share with N trains from the Astoria line). In Manhattan, this route proceeds west under 60th Street (seven blocks north of the 53rd Street route followed by E and F trains), then turns south at Seventh Avenue (two blocks east of ,the E train's Eighth Avenue route and one block 'west of the F train's Sixth Avenue route) to connect with the local tracks of the Broadway BMT line. On the Queens Boulevard line, two local train services are provided - the G and the R. However, G trains turn south to Brooklyn just west of Queens Plaza and do not enter Manhattan. Only R trains enter Manhattan, and the report concerns itself just with them. During the morning peak hour, 12 R trains cross the Queens cordon into Manhattan. They have a functional capacity of 16,800 passengers. Transit Authority counts show current demand of only 12,044 passengers, or 11 percent of total demand at the Queens cordon. This means that R trains have load factors of 72 percent, with sufficient space to accommodate an additional 4,756 passengers. * Load Balancing If enough E and F passengers were willing to make use of this reserve capacity on R trains by transferring at Queens Plaza, overcrowding on the E and F would be significantly reduced, though not eliminated. Under these circumstances, E and F trains would cross the Queens cordon with average load factors of 114 percent (5,556 excess passengers). Only four additional express trains would be needed to eliminate overcrowding, raising the total number of Manhattan-bound trains in the 35 53rd Street tunnel to 31. if these extra trains could be accommodated, E and F trains would cross the Queens cordon with average load factors of 100 percent. It is not clear why Queens Boulevard line passengers do not share the willingness of Flushing line passengers to balance loads between express trains and locals at Queens Plaza in order to reduce overcrowding. After leaving Manhattan, R trains make no additional stops in Queens, make two stops while proceeding west across Manhattan (at Lexington Avenue and at Fifth Avenue), and make four stops after turning south to Herald Square (at 57th Street, 49th Street, 42nd Street, and 34th Street). This totals six stops between Queens Plaza and Herald Square. By contrast, E trains make seven stops between Queens Plaza and Penn Station - at Ely Avenue in Long Island City; at Lexington Avenue, Fifth Avenue, and Seventh Avenue while proceeding west across Manhattan; and at 50th Street, 42nd Street, and 34th Street after turning south on Eighth Avenue. F trains make six stops between Queens Plaza and Herald Square (the same number as the R train) - at Ely Avenue in Long Island City; at Lexington Avenue and Fifth Avenue while proceeding west across Manhattan; and at Rockefeller Center (50th Street), 42nd Street, and 34th Street after turning south on Sixth Avenue. Under-utilization of Queens Boulevard local service into Manhattan has been a fact of life ever since it was initiated in 1954. This suggests that part of the overcrowding problem on the E and F may arise from passenger preferences that need to be studied more carefully. 36 III. T H E H I S T 0 R I C A L C 0 N T E X T Planning decisions often cast long shadows across history. They can shape the lives and habits of subsequent generations, who come to regard them as part of Destiny's own pattern - like the evolution of the stars. This is certainly true of New York's subway system, which is unmatched anywhere in the world for the grandiloquent complexity of its route structure. A system like the Paris Metro may seem equally complex from a glance at its route map. But it is actually as simple as a plate of spaghetti. Each strand is physically separate from all the rest. By contrast, New York's subway system is a bewildering web of interconnected routes that merge and diverge in a subtle, inspired, perverse, even whimsical fashion. None of this is accidental. It is the result of a long series of planning decisions, heavily influenced by the public compromises and private agendas that seemed to make sense at the time. Some of these decisions have impacted the problems we now face in seeking ways to increase subway capacity across the Queens cordon. Since planning for the future cannot escape the realities of the past, it is important that we understand how our subway system came to be the way it is - at least as far as it affects the Queens cordon. THE FLUSHING LINE Construction of the Flushing line was begun during the second decade of the century by a private firm - the Interborough Rapid Transit Company (or IRT). Among other things, it was an outgrowth of a determined attempt to make use of a pair of trolley tunnels that had been built under the East River between midtown Manhattan and Long Island City by another private firm with ambitious goals but inadequate finances. Because of the restricted clearances in these tunnels, any subway line using them would have to operate the narrow and short cars that the IRT company had adopted for its other lines. Hence, the logic of building the Flushing line as an IRT line across 42nd Street, through the trolley tunnels to Long Island City, then eastward on an elevated structure through what was then open country to the little village of Flushing. By providing direct subway service to Midtown Manhattan, it was hoped that the new line would stimulate urban development in the rural areas of Sunnyside, Woodside, Jackson Heights, Elmhurst, and Corona. The real estate boom of the 1920s made this hope a reality. The rural areas developed more rapidly and to a higher density than their original promoters had dreamed possible. The resulting increase in population led to ridership demands on the Flushing 37 line that it was never designed to accommodate. This is one part of the problem that planners face today as the Flushing line's catchment areas experience a new burst of population growth fueled by immigration from Latin America and Asia. It is tempting to consider what today's conditions at the Queens cordon would be like if the original planners had not been mesmerized by the opportunity to make expeditious use of those empty trolley tunnels under the East River. Perhaps the Flushing line would have been built to accommodate the larger subway cars adopted by the Brooklyn-Manhattan Transit Company (the BMT), which was the other private firm that won franchises to build subways in New York. This would have given the Flushing line 16 percent more passenger capacity than it now has. Today's peak hour overcrowding on the Flushing line at the Queens cordon would disappear completely and there would be a six percent capacity reserve to accommodate future growth. Or perhaps the proposal to build the Flushing line would have become mired in planning controversies over whether it made financial sense to spend extra money to enlarge the trolley tunnels - or to ignore them completely and build an entirely new river crossing. Perhaps (like the Second Avenue subway) these controversies might have led to the Flushing line never being built at all, and the area of Queens that it serves would have developed quite differently. This is something we can never know. But the temptation to dream a different history for Queens is irresistible. THE OUEENS-BOULEVARD LINE In retrospect, it is easy to point out mistakes made in planning the Queens Boulevard IND line and blame them for many of the problems we now face at the Queens cordon. But this is too facile to be realistic. It is not at all clear that avoiding these to mistakes" would have made our current planning challenges any easier. In the early 1920s, when the IND system was conceived, no responsible planner could have proceeded on the assumption that Queens would grow as vigorously as it did. More than anything else, this growth is what has caused the overcrowding at the Queens cordon. But planning decisions may have had some impact. To understand how, we must know at least the highlights of the IND's rather unorthodox planning history. The "Independent City-Owned and City-Operated" subway system (the IND's formal name) was essentially the love child of Mayor John Hylan - who was an interesting combination of clubhouse politician, socialist visionary, and vengeful child. Mayor Hylan was passionately committed to giving New Yorkers the best and most comprehensive subway system in the world, which he believed had to be owned and operated by the City. His fulminations against the private "traction interests" make fascinating (sometimes hilarious) 38 reading. Even so, all his frothing at the mouth cannot obscure the intensity of his conviction that public transportation should be a municipal service - with no room for private profit. The route schematic on the next page shows (in highly simplified form) the original concept for the IND system. The heart of it consisted of two north/south trunk line subways on either side of Manhattan. The west side line would run from Washington Heights down Eighth Avenue. The east side line would run from the Bronx down the Grand Concourse and Second Avenue. In lower Manhattan, both lines would cross the East River to Brooklyn. The Eighth Avenue line would run south through Park Slope to a connection with the BMT company's Culver line (which would be taken over by the City) so it could reach Coney Island. The Second Avenue line would run east under Fulton Street to southeast Queens. Finally, a third new Manhattan trunk line would run down Sixth Avenue and connect with the Eighth Avenue line in Greenwich Village. These three trunk lines would be connected in Midtown Manhattan by a new crosstown line that would continue east under the East River and out Queens Boulevard to Jamaica and beyond. This ambitious concept had three clear goals: * To provide through subway service via Manhattan between Brooklyn, Queens, and the Bronx. This would eliminate the need for many Brooklyn residents to pay double fares in order to transfer to between IRT and BMT lines in Manhattan. (Needless to say, Mayor Hylan was a Brooklynite). * To replace the elevated lines in Manhattan on Second, Third, Sixth, and Ninth avenues in order to stimulate commercial redevelopment that would bolster the City's tax rolls. (Mayor Hylan was not above sleeping with anyone who could help him get his way - so long as they weren't too closely associated with the despised private transit companies.) * To overlay, and therefore compete with, the recently completed subway systems of the IRT and BHT companies. These two systems were entirely separate private enterprises with no provision for free transfers between them. Mayor Hylan hoped that the IND would attract riders away from these systems, thereby softening up their companies for acquisition by the City at bargain prices in order to realize his dream of a single publicly owned and operated subway system. (This finally happened 39 Click HERE for graphic. in 1940 - and forty years later, it became apparent that these acquisitions had actually cost the City something less than nothing.) Financial constraints made it impossible to construct this entire system as a single entity. So large portions - including the Second Avenue line in Manhattan - were deferred to "Phase Two of I" leaving the truncated system shown in the route schematic on the next page. The Grand Concourse line was rerouted to connect with the Eighth Avenue line at 145th Street so that the IND would be able to serve all three of New York's main residential boroughs. And a new line connecting downtown Brooklyn with Long Island City was added (more about this in a moment). Even this truncated system was really more than the City could afford to undertake. But Mayor Hylan insisted on proceeding, perhaps assuming that if enough holes were dug and enough streets were disrupted, the necessary funds would somehow be found to complete it. In any case, New Yorkers should be grateful for his willfulness. The necessary funds were eventually found (by Mayor LaGuardia), and the basic IND system as we know it was completed by 1940. Mayor Hylan's emotional involvement with the IND resulted in some personal touches that can most kindly be described as whimsical. For example, the extra southbound platform on the lower level at 42nd Street on the Eighth Avenue line is attributed to him. Its alleged purpose was to prevent the hated IRT company from ever being able to extend its Flushing line west of Eighth Avenue. In the 1960s, the Transit Authority built its brilliantly conceived "Chrystie Street Connection" in Manhattan, which combined the IND and BMT into a single gigantic system that enabled trains from Coney Island to reach the Bronx and Queens via the Sixth Avenue line in Manhattan. The vast range of route options this offered led to the bewildering array of IND/BMT train services we enjoy (or get lost in) today. The magnitude of their astonishing complexity, replete with endless mergings and divergings and track sharings, is best conveyed by the quotation below. It is taken from the 1970 DOT report Simplified Subway Routing: Key To Better Transit In New York City which was prepared by Raleigh D'Adamo, a wise and witty transit maven of the time. Some of the train services have had their letter names and routes changed slightly since it was written. But their essential character is still the same. "The M train merges first with the K train at Myrtle Avenue, secondly with the i at Myrtle Avenue, thirdly with the RR at Court Street, and fourthly with the D and QB trains at DeKalb Avenue. The D merges first with the A at 145th Street, next 40 Click HERE for graphic. with the B at 59th Street, and then - as we have just noted with the m and QB at DeKalb Avenue. The A, as we have already observed, merges with the D at 145th Street, merges next with the E at 42nd Street, then with the E again at Euclid Avenue. The B which, as we have seen, merges -with the D at 59th Street, merges first with the CC at 168th Street and later with the N at Dekalb Avenue. The RR, which we stated merges with the N at Court Street, merges with the ZE at Lexington Avenue - the EE having first merged with the GG at Continental Avenue. The E which, as we said, merges with the A at 42nd Street, will have first merged with the F at Continental Avenue. The F, after thus merging with the B at Continental Avenue, later merges with the K at 50th Street. While the K, after merging with the F at 50th Street, merges with the J and the N at Essex Street - which is where we started" has some planning logic to support it. For whatever reason, the Queens Boulevard line was built with no access to Manhattan for its local trains. This forced Manhattan Bound local riders to complete their trips, which established the sardine-can tradition of jammed express trains in the 53rd Street tunnel. In 1954, this apparent planning error was corrected when a new connection opened between the Queens Boulevard line's local tracks and the BMT's 60th Street tunnel into Manhattan. The original concept of a four track IND subway crossing the Queens cordon was rescued from the scrap heap to ease the overcrowding on express trains using the 53rd Street tunnel. But riders largely ignored the new service option - and still do today. As we saw in the preceding section, peak hour R trains from the Queens Boulevard line's local tracks regularly depart from Queens Plaza for the 60th Street tunnel with load factors of only 72 percent. Passengers insist on jamming their way into crushloaded E and F trains, which depart Queens Plaza for Manhattan with average load factors of 127 percent. If riders could be persuaded to take advantage of the unused capacity on R trains, load factors on the E and F would be reduced to 115 percent - still overcrowded, but more tolerably so. The reasons behind this passenger behavior are unclear. But they need to be understood. Do they simply reflect planning errors that have remained obscure for the past forty years? Or do they involve subtler aspects of rider psychology that might dictate a very different approach to addressing the overcrowding problems at the Queens cordon? During the 1960s, the Transit Authority began serious planning for a long-contemplated new subway line to serve central Queens. In its final form, this line would branch off the Queens Boulevard line just west of Continental Avenue, run generally south for a few blocks under Yellowstone Boulevard to the LIRR main line, then come to the surface and run generally west next to the LIRR main line to Sunnyside Yard. From there, it would run underground to a new station in Queens Plaza (which would provide a passenger transfer to the Queens Boulevard line's express station), then run generally west under 41st Avenue and through a new tunnel under the East River to 63rd Street in Manhattan. It would continue west under 63rd Street to Sixth Avenue (and a connection with the Sixth Avenue IND line) and to Seventh Avenue (and a connection with the express tracks of the Broadway BMT line). Between Sunnyside Yard and Manhattan, this line would also provide a separate lower level to accommodate LIRR trains. This would give the LIRR a second means of access to Manhattan, enabling 42 some of its trains to run into Grand Central Terminal. Because of funding shortages, the Queens portion of the line was never built. But the very expensive Manhattan portions and the new river tunnel (including the lower level for the LIRR) were completed by the 1980's - at costs that today seem like bargain prices. Subway service now operates from the Sixth Avenue line in Manhattan to a temporary terminal in Long Island City at 41st Street and 21st Avenue. The MTA proposes to connect this line to the Queens Boulevard line just east of the Queens Plaza express station so that through service between central Queens and Manhattan via the 63rd Street line can be operated. 43 I V. S 0 L U T I 0 N 0 P T I 0 N S The only practical way to eliminate subway overcrowding at the Queens cordon is to increase subway capacity. The options for doing so can be grouped under three categories. Run longer trains. Connect existing subway lines in ways that allow more trains to be run. Build new subway lines. Many proposals, have been advanced over the years for increasing public transportation capacity between Queens and Manhattan. The options discussed below reflect proposals that share two common characteristics. First, they involve only the subway system. Therefore, proposals involving LIRR improvements, expanded Amtrak or Metro North service via the Hell Gate Bridge, and development of new systems involving trolley (light rail) or people mover technology are outside the scope of this report. Second, they are not mutually exclusive. Implementing one does not preclude implementing any of the others at a later date. This makes it possible to phase in various sequences of options over a long period of time, providing incremental increases in capacity to match growth in demand. An important factor in evaluating these options is the proposal to build a transfer station for Amtrak and LIRR trains in Sunnyside yard, possibly at the Queens Boulevard bridge. It would be highly desirable for this station to have convenient transfers to subway lines in order to enhance Long Island City's accessibility by public transportation. This factor is considered in the discussion of each option. Some people believe that measures to increase subway capacity in the Central Queens Corridor should also incorporate provisions for serving JFK and LaGuardia airports. However, the special characteristics and needs of air passengers suggest that it is neither feasible nor desirable to attempt to serve them by subway. Therefore, airport access is not a factor that is considered in discussing each option. 44 RUN LONGER TRAINS In theory, the simplest way to increase subway capacity is to run longer trains on services that experience peak hour overcrowding. The table on the next page, And the diagram that follows it, show the impact of running longer trains in the Central Queens Corridor. This approach has two notable advantages. * New capacity is closely targeted to those train services that have excess demand. Therefore, the new capacity can be fully utilized without requiring passengers to change their travel patterns. However, passengers must distribute themselves more or less evenly throughout each train if this approach is to eliminate overcrowding. * New capacity is achieved with minimal increases in subway operating costs. There is no change in the number of trains being operated, the number of train crews required, and the number of other subway operating personnel needed. Each train is simply able to carry more passengers. Operating cost increases are limited to the additional electricity consumed by the extra cars, the maintenance of these extra cars, and the somewhat greater track wear caused by running more cars. Longer trains require two kinds of modifications to the subway lines on which they operate. * Station platforms must be extended. This is usually (but not always) a relatively simple and only moderately costly undertaking. * Signal systems must be replaced. Existing subway signal systems are based on the fixed-length block principle. The length of these blocks determines how long trains can be. When blocks are too short to allow operation of longer trains, the signal systems on the lines where these trains are to operate must be replaced. This is a complex and expensive undertaking. But important savings can be possible if signal system modification dovetails with the need to replace signals that have reached the end of their economically useful life. 45 Click HERE for graphic. Click HERE for graphic. Click HERE for graphic. During the 1950's and 60's, the Transit Authority used this approach to increase capacity on many portions of the IRT and BMT systems. 1. The Flushing Line As noted in Section II, the Flushing line currently operates trains of eleven cars that are 550 feet long and have a functional capacity of 1,210 passengers. During peak periods, these trains have load factors of 108 percent at the cordon crossing. This overcrowding could be eliminated by increasing train lengths to twelve cars (making them 600 feet long). Trains would then be crossing the cordon with load factors of 99 percent. If a reserve capacity of ten percent was desired to allow for future growth, trains of fourteen cars (700 feet long) would be needed. With a ten percent increase in demand, these trains would cross the cordon with load factors of 94 percent if this increase in demand materialized. Unlike most other New York City subway lines, the Flushing line is a "discrete" line. That is, its trains operate only on this line and do not share tracks with other train services. This simplifies the task of converting the Flushing line to accommodate longer trains. Station platform extensions and signal replacements would be limited only to this line. Additional cars would also have to be purchased, increasing the size of the Flushing line car fleet by 27 percent. Storage tracks and maintenance facilities in the Flushing line's yard next to Shea Stadium would have to be modified to accommodate longer trains. Since the Flushing line crosses Sunnyside yard on an elevated structure above the Queens Boulevard bridge, it is already well situated for transfers to and from a new Amtrak/LIRR station at this location. 2. The Oueens Boulevard Line E and F trains on the Queens Boulevard line currently operate eight car (Class R44 and later) trains that are 600 feet long. Each train has a functional capacity of 1,400 passengers. Excess demand causes E trains to have load factors of 122 percent, and F trains to have load factors of 133 percent, at the cordon crossing. 46 This overcrowding could be eliminated by operating ten car E trains (750 feet long) and eleven car F trains (825 feet long). This would enable E trains to cross the cordon with load factors of 98 percent and F trains to cross the cordon with load factors of 96 percent. If a ten percent reserve capacity was desired to allow for future growth, E trains would have to be 11 cars (825 feet) long and F trains would have to be 12 cars (900 feet) long. Under these demand conditions, E trains would cross the cordon with load factors of 98 percent. F trains would have load factors of 97 percent. Express station platforms on the Queens Boulevard and 53rd Street lines would have to be lengthened to 900 feet and the signal system on the express tracks would have to be replaced to accommodate fourteen car trains. Similar modifications would also have to be made on the other lines that E and F trains serve. * E trains operate on the Eighth Avenue line in Manhattan between 50th Street and the stub end station at the World Trade Center. * F trains operate on the Sixth Avenue line in Manhattan and the Culver line in Brooklyn. Therefore, modifications to allow longer trains would be more extensive in scope than for the Flushing line. The car fleets serving E and F trains would have to be increased in size by 75 percent to enable longer trains to be run. Storage tracks and maintenance facilities in the Jamaica and Culver (Coney Island) yards would have to be modified to accommodate longer trains. The Queens Boulevard line's Queens Plaza station is roughly 1000 feet from the proposed location of a new Amtrak/LIRR station in Sunnyside yard. Anything done to enhance transfer opportunities between the two stations would benefit from greater capacity on E and F trains. 3. The Astoria Line N trains serving the Astoria line have the same functional capacity as E and F trains, or 1,400 passengers each. During the morning peak hour, they cross the cordon into Manhattan with load factors of 12 47 percent. To eliminate this overcrowding, one additional car would have to be added to each N train, making it nine cars (675 feet) long. This would enable N trains to cross the cordon with load factors of 99 percent. A ten percent reserve capacity would require ten car trains that would be 750 long. Under these demand conditions, N trains would cross the cordon with load factors of 98 percent. Station platforms on the Astoria line would have to be lengthened to 750 feet and the signal system would have to be replaced. Once all four tracks on the Manhattan Bridge are available to subway trains, N trains will operate on the express tracks of the Broadway BMT, the west tracks of the Manhattan Bridge, and the Fourth Avenue and Sea Beach lines in Brooklyn. Station platform and signal system modifications would be needed on these lines to accommodate longer trains. The car fleet serving the N train would have to be increased by 25 percent. Storage tracks and maintenance facilities in the Coney Island yard would have to be modified to accommodate longer trains. The Queensboro Plaza station that is used by N trains may be too far from the proposed Amtrak/LIRR station in Sunnyside yard to allow convenient transfers unless additional facilities are provided (such as moving sidewalks or shuttle transit services). Increasing capacity on N trains would enhance the benefits from these facilities. The appeal of the longer trains option is its apparent simplicity, highly targeted nature, and minimal impact on operating costs. However, passengers would have to distribute themselves more evenly throughout each train than they do now if this new capacity is to eliminate overcrowding. Also, the implications of running trains whose length could be up to three times longer than a football field must be evaluated carefully. For example, it is possible that two conductors per train would be needed during commuting periods to assure adequate passenger safety at high demand stations. NEW CONNECTIONS BETWEEN EXISTING SUBWAY LINES Since the 1930's when the Queens Boulevard line was completed, two major modifications to the subway system were undertaken to provide additional passenger capacity into the Manhattan CBD. One (mentioned above) involved changes that allowed longer trains to be run on IRT and BMT lines. The second, which opened for service in 48 1968, was the Chrystie Street connection that connected the Sixth Avenue line in Manhattan to the east tracks on the Manhattan Bridge. This unified the IND and BMT into a single system and allowed more trains to be run. The existence of the 63rd Street line, which extends from the Sixth Avenue IND and Broadway BMT lines in Manhattan and now terminates at 21st Avenue in Long Island City, provides new opportunities for various connections that would increase passenger capacity in Queens. However, unlike options that involve running longer trains, Queens subway riders would have alter their current travel patterns if this new capacity is to eliminate overcrowded trains. The extent to which they would do so remains a matter of speculation. This is not intended as an argument against building new connections. It is mentioned only to stress the importance of careful analysis in assessing their benefits. 1. The Oueens Boulevard Line Connection The MTA has proposed to connect the upper level tracks of the 63rd Street line to the local and express tracks of the Queens Boulevard line at a point just east of Queens Plaza. One major advantage of this connection is the operating flexibility it would provide when service disruptions occur. This may seem mundane, but it can be very important. Disruptions are inevitable on a huge system running many trains, even when a high degree of reliability is achieved. Connecting the 63rd Street line to the Queens Boulevard line would give E, F, and R trains a second route to and from Manhattan when their main route is temporarily blocked. Without an alternative route, service comes to a halt when a disruption occurs. Eventually, service in the other direction is affected because trains are unable to reach their terminals and reverse. The recovery from such a disruption can be a lengthy process. The existence of an alternative route allows trains to bypass the disruption and continue to their terminals, where they can reverse and resume their regular routes in the other direction. Recovery periods are much shorter when this can be done. This has nothing to do with passenger capacity as such. But it helps safeguard daily service reliability, which is no small benefit. For that reason, some sort of connection to the Queens Boulevard line is worth serious consideration regardless of which option seems best suited to providing more passenger capacity. 49 From the perspective of new passenger capacity, this connection would enable the Queens Boulevard line to operate up to 14 additional Manhattan trains per hour in each direction during peak demand periods. These new trains would utilize half the track.capacity of the 63rd Street line's upper level. They would increase the Queens Boulevard line's Manhattan-bound functional capacity by 19,600 passengers (34 percent) during the morning peak hour. However, assessing the functional benefits of this new capacity is a bit tricky. The 14 additional trains would operate on the Queens Boulevard line's local tracks, where they would replace G trains that do not now enter Manhattan. As noted in Section II, passengers have demonstrated a reluctance to make use of the capacity that already exists on Queens Boulevard line local trains. Currently, R locals cross the cordon into Manhattan during the morning peak hour with load factors of only 72 percent - even while E and F express trains have load factors of 122 percent and 133 percent respectively. Therefore, it is reasonable to speculate about the benefits to be gained by increasing the number of already under-utilized local trains. The MTA has been running simulations of various service patterns in an effort to explore possibilities options for making the best use of the new track capacity. The table on the next page and the diagrams that follow it show the details of one of the most promising simulations ("Service Option 4"). * E trains would become locals on the Queens Boulevard line and continue their present routings through the 53rd Street tunnel and down Eighth Avenue. Under current demand conditions, they would cross the cordon into Manhattan during the morning peak hour with load factors of 79 percent (down from their present 122 percent). If demand increased by 10 percent, their load factors would be 87 percent at the cordon crossing (down from 135 percent under existing service patterns). * F trains would continue their present routings, running as expresses on the Queens Boulevard line, through the 53rd Street tunnel, and down the Sixth Avenue line. Under current demand conditions, they would cross the cordon into Manhattan with load factors of 100 percent (down from their present 133 percent). If demand increased by 10 percent, they would cross the cordon with load factors of 105 50 Click HERE for graphic. Click HERE for graphic. Click HERE for graphic. percent (down from 146 percent under current service patterns). * A new service ("V trains") would replaced E trains on the Queens Boulevard line's express tracks. It would enter Manhattan through the 63rd Street tunnel and run down the Sixth Avenue line. Under current demand conditions, the MTA's simulations show that V trains would cross the cordon with load factors of 110 percent (reflecting apparent passenger preferences for using expresses, even when they are overcrowded). If demand increased by 10 percent, they would cross the cordon with load factors of 121 percent. * R trains would continue their present routings, running as locals on the Queens Boulevard line, entering Manhattan via the 60th Street tunnel, and running down the Broadway BMT line. Under current demand conditions, they would cross the cordon into Manhattan during the morning peak hour with load factors of 65 percent (down from their present 72 percent). If demand increased by 10 percent, they would cross the cordon with load factors of 71 percent (down from 79 percent under existing service patterns). In summary, the simulation shows that the number of Queens Boulevard line trains entering Manhattan during the morning peak hour would rise by 36 percent, from 39 to 53. This represents an increase in functional capacity of 19,600 passengers per hour, from 54,600 per hour to 74,200. Current Manhattan-bound demand totals 60,156 passengers during the morning peak hour. If passengers allocated themselves among these 53 trains in ways that balanced loads (which is not what they currently do), trains would cross the cordon with average load factors of 88 percent. This would leave a 12 percent capacity reserve to accommodate future growth. To achieve this theoretical optimum, ways would have to be found to induce passengers to balance trains loads more rationally (as they already appear to do between Flushing line trains and Astoria line N trains). The MTA's Service Option 4 simulation also shows some benefits for the Flushing line. It suggests that 11 percent of line's passengers would switch to Queens 51 Boulevard line trains - either at the Roosevelt Avenue transfer station or by entering the subway system at Queens Boulevard line stations rather than Flushing line stations. Under current peak hour demand conditions, Flushing line trains would cross the cordon into Manhattan with load factors of 97 percent of capacity (down from the present 108 percent). If demand increased by 10 percent, these trains would cross the cordon with load factors of 106 seven percent (down from 119 percent under existing service patterns). At the same time, N trains would cross the cordon in Manhattan with load factors of 104 percent (down from 112 percent). With 10 percent more demand, they would cross the cordon with load factors of 114 percent (down from 123 percent under existing service patterns). Therefore, with the Flushing line included, the simulation shows the following picture at the Queens cordon: * The number of Manhattan-bound trains during the morning peak hour would increase by 18 percent, from 77 to 91. This would increase functional capacity by 19 percent (19,600 passengers), from 102,670 passengers per hour to 122,270. * Given current peak hour demand of 112,790 passengers, the average train would have a load factor of 92 percent of capacity if passengers allocated themselves among trains to balance loads. This would leave an eight percent capacity reserve for future growth. * In the absence of optimal load balancing, the simulation shows the following load factors for trains crossing the Queens cordon: - 97 percent on Flushing line trains under current demand conditions, rising to 106 percent if demand increases by 10 percent. - 104 percent on Astoria N trains under current demand conditions, rising to 114 percent with a 10 percent increase in demand. - 83 percent on Queens Boulevard E locals under current demand conditions, rising to 87 percent with a 10 percent increase in demand. 52 - 110 percent on Queens Boulevard V expresses under current demand conditions, rising to 121 percent with a 10 percent increase in demand. - 95 percent on Queens Boulevard F expresses under current demand conditions, rising to 105 percent if demand increases by 10 percent. - 65 percent on Queens Boulevard R locals under current demand conditions, rising to 71 percent if demand increases by 10 percent. It seems apparent that rational load balancing among trains in the Central Queens Corridor is the key factor in enabling this option to eliminate subway overcrowding by itself. Further simulations may identify service patterns that would enhance the likelihood of this happening. Since the connection with the Queens Boulevard line would be made east of the Queens Plaza station, there would be no transfer opportunities between 63rd Street trains and the proposed Amtrak/LIRR station in Sunnyside yard. 3. Reverse Signaling For The Oueens Boulevard Line The MTA has discussed this option as a "Phase Two" follow-on to the Queens Boulevard line connection. it involves re- signaling the line's two express tracks so that trains could run in either direction. This would permit a "3 & 1" operating configuration during high demand periods. Both express tracks and one local track would run Manhattan-bound trains during the morning commuting period, and just the reverse in the afternoon. Such a configuration would double the number of Manhattan- bound express trains during the morning peak hour - from 27 to 54 - and would utilize 100 percent of the track capacity on the 63rd Street line. One service pattern that could take advantage of this is illustrated in the diagram on the next page (though details are not available on how passengers would allocate themselves among these trains). This pattern is outlined below. * 14 E expresses per hour operating through the 53rd Street tunnel and down Eighth Avenue (the same as now). 53 Click HERE for graphic. * 13 F expresses per hour operating through the 53rd Street tunnel and down Sixth Avenue (the same as now). * 13 V expresses per hour operating through the 63rd Street tunnel and down Sixth Avenue (the same as with the Queens Boulevard connection line alone). * 14 Q expresses per hour operating through the 63rd Street tunnel and down Broadway (on the express tracks). * 12 R locals per hour operating through the 60th Street tunnel and down the Broadway local tracks (the same as now, but 13 fewer than with Service option 1 for the Queens Boulevard line connection alone). In other words, the Queens Boulevard line's total functional capacity Manhattan bound (expresses and locals) would be 92,400 passengers per hour. This is 70 percent more than now and 27 percent more than with the Queens Boulevard line connection alone. Under current demand conditions - and assuming the same shift of Flushing line passengers to the Queens Boulevard line as under the Queens Boulevard line connection alone - the average Manhattan-bound Queens Boulevard train (local and express) would cross the cordon with a load factor of 70 percent, leaving a 30 percent capacity reserve for future growth. Further analysis is needed to estimate how total demand would allocate itself among the four train services. But this capacity increase could mean significantly higher operating costs for the Queens Boulevard line. During the morning commuting period, only one track would be available for trains running towards Jamaica. This might be insufficient to accommodate all of the trains that would have to reach the two Jamaica terminals in order to fully stock the three tracks operating to Manhattan. In such a case, a number of "single-trip" trains (and their crews) would have to be put in service at the Jamaica terminals, which would increase operating costs. The same problem could occur in the opposite direction during the afternoon commuting period. Further analysis is needed to determine the magnitude of these operating cost increases. 54 Some people have argued that 3 & 1 operation on the Queens Boulevard line could be a dicey undertaking because it allows so little room for error. Each train must be in the right place at the right time to avoid delays that could snowball into major disruptions during commuting periods. The validity of this argument needs to be evaluated carefully in order to determine whether the benefits of 3 & 1 operation might turn out to be more apparent than real. As with the Queens Boulevard connection alone, this option offers no transfer opportunities between 63rd Street trains and the proposed Amtrak/LIRR station in Sunnyside yard. 4. Connecting The Flushing Line It would be possible to connect the Flushing line to the 63rd'-Street line and have its trains run on the Sixth Avenue IND and Broadway BMT lines in Manhattan. This would require that the Flushing line be converted to IND standards so that it could operate longer trains composed of wider cars. Doing so would increase the functional capacity of each Flushing line train by 16 percent, from 1,210 passengers to 1,400 passengers. The result would give the Flushing line a functional capacity of 37,800 passengers per hour in each direction vs. the present 32,670 passengers per hour. The diagram on the next page shows the impact of this option. Under current demand conditions, this conversion would mean that Flushing line trains would cross the cordon into Manhattan during the morning peak hour with load factors of 94 percent (down from the present 108 percent). A six percent capacity reserve would be available for future growth or to help relieve overcrowding on E, F, and N trains. If Flushing line trains absorbed the current N train overload through transfers at the Queensboro Plaza station, its capacity reserve would shrink to one percent. A reserve this small would leave little