The Limits of Moderate-Capacity Transit

Gas prices in the first nine months of 2006 were at their highest levels (after adjusting for inflation) in twenty-five years. Most transit agencies made the most of this, with some gaining huge increases in ridership over the first nine months of previous year.

  • Flagstaff saw a 49-percent increase in bus ridership;
  • Tucsan saw a 26-percent increase;
  • Colorado Springs gained 22-percent more riders;
  • Tulsa 20 percent;
  • Santa Fe got 15-percent more riders.

It enhances a viagra from india woman’s sexual desire as compared to women. NS5B non-nucleoside polymerase order viagra no prescription inhibitors (NNPIs)- block the action the HCV NS5A protein and interfere with making new viruses. Pfizer really has its job cut viagra buy uk out due to the health problem, male organ does not become hard enough for penetration and causing erection to turn an intercourse into sexual pleasure. Here, some of the factors have been mentioned: cialis generico canada FDA Approved Sildenafil ED Drug The key ingredient is finasteride, a medicine that acts as an inhibitor.
While these are exceptional, APTA reports that American transit systems carried 3 percent more transit riders than in 2005. So why did the transit system in Portland, the city that supposedly loves transit, actually carry less riders in 2006 than the year before?

Moderate-capacity transit in Portland.

The answer has to do with Portland’s choice of technology. Though light rail is often called “high-capacity transit,” in fact it is anything but. After all, the “light” in light rail refers to light loads and is a comparison with subways or elevateds, also known as heavy rail, meaning they can carry heavy loads. Light rail’s load limits have become a particular problem for Portland.

Despite claims that a light-rail line can move as many people as an eight-lane freeway, light rail’s capacity is in some cases less than a single freeway lane. A system’s capacity to move people depends on vehicle frequencies or headways, the number of occupants per vehicle, and vehicle speeds.

Freeway lanes can move about 1,800 cars per hour at speed, or one every two seconds. Exclusive bus lanes can move about 360 buses per hour, or one every ten seconds. Rail frequencies are much lower and depend on the quality of signaling. Typically, light-rail systems allow no more than one train every three minues, while some heavy-rail systems may allow one train every two minutes.

Many rail advocates compare rail capacities with average auto occupancies, but this is apples to oranges. Auto capacities depend on the size of auto, but let’s assume an average of five. Bus capacities are typically about 80 people including standees.

Train capacities depend on the number of cars per train. Heavy rail runs exclusively on its own rights of way, so the number of cars is limited by the length of station platforms. Most heavy-rail lines are built for trains of eight cars. Subway cars can carry about 180 people, so a train can carry more than 1,400 people.

In contrast, most light-rail trains spend at least part of their time on city streets. To avoid blocking traffic, this means that train length is limited by the shortest blocks where a train might stop. Downtown Portland has unusually small blocks — only 200 feet on a side — and because light-rail cars are about 90-feet long, Portland can run no more than two cars at once. Many other light-rail cities have three-car trains, but only a few run more than three cars per train..

The capacity of a light-rail car is about 170 people, so Portland trains can carry no more than 340 people. All these transit capacities, by the way, are “crush capacities,” meaning Tokyo-like crowding rarely seen in the U.S. I rode a very crowded DC subway today and counted 80 seated and 50 standing, about 50 short of the crush capacity.

Many rail advocates ignore speed, but speed is an important component of the amount of work a transportation system does. On the 50th anniversery of the Golden Gate Bridge, the bridge was briefly closed to autos and opened excusively to pedestrians. More people walked across the bridge in one hour than had ever driven across it in that amount of time. But no one thinks the bridge would do more work if it were closed to autos because autos are so much faster than walking.

Light-rail trains average about 22 mph. Some heavy-rail lines average 40 mph. Most urban freeways can handle cars or buses at 65 mph.

Multiply these numbers out and we get the following capacities in passenger miles per hour:

2-car LRT          149,600
3-car LRT          244,400
8-car HRT        1,728,000
Freeway lane       585,000
Bus lane         1,872,000

Even a four-car light-rail line could do less work than a freeway lane open to autos. A freeway lane open only to buses could do a little more work than a heavy-rail line and more than eight times as much work as a three-car light-rail line. Even if bus-lane speeds averaged only 10 mph (as on normal city arterials), the bus lane could do far more work than any light-rail line.

All of these numbers, of course, are far more than actual use. In 2005, the average auto had 1.6 people; the average transit bus carried 10.3 people; the average light-rail car carried 25.5 people; and the average heavy-rail car carried 22.9 people. Based on these numbers, typical hourly flows are shown in the table below.

2-car LRT           22,176
3-car LRT           33,274
8-car HRT          219,840
Freeway lane       187,200
Bus lane           241,020

Once again, the bus lanes beat even heavy rail and carry seven times as many people as three-car light rail. Buses have another advantage: individual buses can easily serve numerous neighborhoods or districts, then get on the bus lanes and head for downtown or other major centers. By comparison, rail cars tend to empty out as they near urban fringes. That means the buses can carry more people over a longer portion of their routes than rail lines.

Actual use is more realistic than capacity, but sometimes capacity is important. Portland’s light-rail system is more successful than most, but this very success works against it because of the system’s limited capacity.

In deciding to build light rail, Portland selected the wrong technology for a city with short blocks. Even a city with long blocks would do better by building virtual exclusive bus lanes — bus lanes open to toll-paying cars with tolls set to prevent any congestion — rather than light rail.

Now Portland is making matters even worse by tearing up its bus mall to add a new light-rail line to it. The bus mall is already near to its capacity for moving people. But adding light rail will actually reduce the route’s capacity for moving people.

Tearing up the bus mall to install light rail. Photo courtesy of PDX Pete.

Why not put the light rail on some street parallel to the bus mall? Because downtown businesses all agreed that, while they wanted light rail, they did not want it on their streets. Construction of the bus mall in the 1970s put many stores out of business, and retailers and restaurants on other streets did not want the same to happen to them. Since most of the businesses were already gone from the bus mall, that was the only place to build.

So don’t expect Portland transit ridership to grow much any time soon. Just remember that in Portland, reality doesn’t matter; image does.

Bookmark the permalink.

About The Antiplanner

The Antiplanner is a forester and economist with more than fifty years of experience critiquing government land-use and transportation plans.

7 Responses to The Limits of Moderate-Capacity Transit

  1. msetty says:

    For my detailed reply to The Antiplanner’s latest transit screed above, see my website, http://www.publictransit.us.

  2. msetty says:

    As for the very slight 0.09% drop in TriMet patronage in 2006, this is due to cuts in non-productive bus service. It’s interesting that, while they also reduced rail service somewhat, overall rail patronage did increase somewhat, around 2.66% for the year in the data cited by Randal. This also means total passenger miles carried by TriMet increased overall slightly, since rail trips are almost twice as long as bus trips.

  3. johngalt says:

    her post is here:

    In today’s post at his website, The Antiplanner (www.ti.org/antiplanner) apparently thinks that he’s stumbled onto some profound findings “proving” why automobiles and buses are “better” alternatives than any form of light rail transit.
    The Antiplanner multiplies average capacities per train or vehicle by his assumptions regarding average speed, and then claims the resulting numbers are the “capacities” of each mode per hour. These calculations are nominally correct, and represent the theoretical maximum number of passenger miles that could be produced by the total fleet of various kinds of vehicles that could travel past a given point in an hour. Thus, for 2-car light rail trains, this is 6,800 people per hour, with 20 trains per hour times the average speed of Portland MAX, or 149,600 passenger miles per hour.
    For automobiles, this is 1,800 automobiles per hour times 4 persons per auto, times an assumed 65 mph, or 585,000 passenger miles per hour. According to this Antiplanner “calculus” the fleet of buses required to maximize theoretical capacity consists of about 600 vehicles, the maximum rush hour bus “pullout” of Portland’s TriMet for its entire service area.
    The transition from theoretical calculations to the “real world” is not a straight line, contrary to what The Antiplanner is implying.
    The capacity of a particular form of transportation technology has relatively little to do with speed, but everything to do with the minimum frequency that trains can operate safely and reliably past a given point, multiplied by the maximum and the maximum loadings per vehicle that is acceptable to transit passengers, or by maximum number of automobiles times average load. For surface transit that must contend with signalized intersections, capacity and reliability dictate minimum headways of at least three minutes. In exceptional cases with a very highly disciplined transit operation, this minimum frequency on any given route is every 90 seconds, as experienced on a combined surface LRT/bus lanes in downtown Calgary.
    This has been found to be the case for both buses and surface light rail. In Los Angeles, surface Metro Rapid buses operating more frequently than every three minutes (e.g., two signal cycles) proved to provide unreliable schedules, e.g., buses would “bunch,” some buses would lose time unnecessarily at traffic signals, and so forth. This also limits reliable frequencies on any one bus line to every 3 minutes, though multiple bus routes on a surface BRT route or busway will inevitably “bunch” at traffic signals.
    The Antiplanner is nominally correct that a freeway lane can carry around 600 buses per hour; however, this rate is actually achieved only in New York City and then only because the Port Authority Bus Terminal has nearly 300 bus bays that can absorb such volumes. In Portland, the 5th and 6th Street transit mall was accommodating up to 200 buses per hour in each direction with multiple bus stops and a travel lane that effectively means two bus lanes in each direction. In Ottawa, the busway system dumps similar volumes of buses onto two main downtown streets, and is very close to its practical capacity. At typical average loads per bus of 25 to 35 passengers (busy routes more, lightly patronized routes less), surface bus capacity is comparable to surface street LRT such as MAX.
    In contrast, in Portland the practical capacity of 2-car LRT trains operating on 3-minute headways is 5,400 passengers per hour–assuming a large enough fleet of LRT vehicles to operate this level of service in the first place. At headways of every 90 seconds, which is quite feasible and reliable on private right-of-way but problematic to on-street running, this capacity is doubled.
    I am using 135 persons per MAX LRT car here, or 270 per train, on the basis that such loads per vehicle are regularly experienced on MAX light rail, this figure is the TriMet maximum loading standard, this appears to be the maximum acceptable load to TriMet riders, and it also is consistent with findings by Mr. Demery.
    When the TriMet MAX line along I-205 opens in 2009, the MAX East Line along the Banfield Freeway will require more frequent trains than every 3 minutes (assuming the demand estimates are accurate!) This will exceed the reliable carrying capacity of the current MAX street trackage in downtown Portland. TriMet’s strategy is to double MAX capacity through downtown by constructing additional LRT tracks on the Transit Mall. This will be near maximum MAX track capacity through the Lloyd Center area, but should be sufficient to provide reliable operation in the downtown core. To ensure reliable service across the Steel Bridge, it would make sense to add 2 additional tracks–4 across the bridge–between the junction with the Yellow Line (North Interstate Avenue), and the divergence between the existing LRT line and the new Mall trackage.
    Coming back to comparing capacities of a transit lane versus a freeway lane in the Portland context, MAX light rail capacities at 3-minute headways with 2 car trains is 5,400 passengers per hour past a given point; at 90-second headways with 2 car trains is 10,800 passengers per hour; with 1,800 automobiles per hour at 1.1 persons per vehicle, capacity is slightly less than 2,000 persons per hour.
    In the case of both the Sunset Highway and Banfield Freeway corridors, the parallel MAX lines are carrying slightly more than one freeway lane during the peak hour, e.g., 2,300-2,500 passengers. With the addition of the I-205 line in 2009, the MAX Banfield line will be carrying 4,000-5,000 during the peak hour, roughly 2 lanes of freeway traffic.
    Both the Sunset Highway and Banfield Freeway corridors are limited to six freeway lanes as a practical matter. Even adding one lane in each direction cold easily exceed $50 million per lane mile, mainly due to the very large number of structures and difficult geographic conditions. As travel demand in these corridors grow, it will be much more cost-effective to add cars to the MAX light rail vehicle fleet than spend additional billions on highway capacity improvements particularly in the key stretches near downtown Portland.
    This is also the logic behind including a $500 million LRT line as part of the $2 billion proposed new I-5 crossing of the Columbia River–giving the I-5 corridor massive reserve capacity well beyond that of a 6- to 8-lane freeway. The proposed Vancouver MAX route would provide about 25% of the peak hour capacity across the river, commensurate with LRT’s share of the overall project cost–and avoiding the additional negative impacts of otherwise needed freeway capacity improvements through North Portland.
    I estimate that operating the equivalent capacity of buses in the Sunset Highway and Banfield corridors to match existing and future MAX train capacity would require the equivalent of 2 or 3 additional bus malls, assuming sufficient downtown capacity is still made available for buses from the rest of Portland.
    As The Antiplanner regularly points out, it is is true that a typical freeway lane under current U.S. conditions carries 2-3 times as much daily traffic as a typical LRT lane. However, as The Antiplanner also knows, current U.S. conditions include chronic underpricing of automobile travel in urban areas and thus dramatically increasing congestion, the pernicious impacts of $300 billion+ per year in “free parking” as documented by Dr. Donald Shoup of UCLA, as well as the yet-to-be-calculated motor vehicle contribution to global warming (which I believe is “real”–but The Antiplanner is highly skeptical of).

  4. msetty says:

    I put my retort to Randal on my website because I didn’t want to leave comments at TWO locations. I suppose you don’t want to build traffic at MY website, what little there is??

  5. msetty says:

    I put my retort to Randal on my website because I didn’t want to have to develop comments at TWO locations. I suppose you don’t want to build traffic at MY website, what little there is??

  6. Pingback: » The Antiplanner

  7. Pingback: Peak Tyranny » The Antiplanner

Leave a Reply