Quito, Ecuador:
El Trole (“The Trolleybus”)
a www.publictransit.us special webreport
 
The city of San Francisco de Quito, capital of Ecuador, stands in a narrow valley high in the Andes (average elevation 2,850 meters (9,300 feet) above sea level). Known for the Spanish Baroque architecture in its 500-year-old centro histórico (“Old Town”), Quito was one of the first two cities designated by UNESCO as a “World Heritage Cultural Site” (in 1979; the other was Praha, better known to English speakers as Prague). The current population of the Municipalidad del Distrito Metropolitano de Quito (MDMQ) is about 1.4 million.
Quito opened a north-south busway worked by articulated trolleybuses in 1996. Officially the Sistema Trolebús del Distrito Metropolitano de Quito, this facility is known among Quiteños as El Trole. It quickly attracted very heavy traffic, reported at more than 200,000 per weekday. The line was extended in 2000, although subsequent busways use diesel-powered vehicles.
We found many interesting operational and technical details on the official website: http://www.trolebus.gov.ec (in Spanish). This was the “primary source” for the information presented below.
 
Description:
The current line length is stated at 16.1 km (= 10.0 miles). El Trole serves two radial corridors, extending roughly 6 km (4 mi) north and 10 km (6 mi) south from the colonial centro histórico. Northern Quito is described as the city’s “modern” high-rise area, with financial center, various businesses and shopping centers. Southern Quito is described as a working-class residential area.
Trolleybuses operate in curb lanes on one-way streets through the centro histórico. The remainder of the line uses dedicated median lanes along major streets.
The line has 39 paradas (“stops”) between terminals. These are patterned after the “tube stations” in Curitiba, Brazil. Vehicles are built for platform-level loading, and have three double-width doors. Bus-mounted plates fold down to cover the gap between vehicle and platform. Passengers pay fares and line up inside the enclosures at the paradas, climbing a few stairs up to “platform level.” Platforms are built long enough to accommodate one vehicle at a time (per direction). This, in short, is a simple, economical and workable arrangement.
The average distance between stations is stated at 400 meters (= 0.2 mi).
A description with photos is available online: http://www.tramz.com/ec/q/00.html A map of the system is here: http://www.tramz.com/ec/q/map.html These are among the excellent pages on public transport in Latin America maintained by Allen Morrison.
(A recent addition to Morrison’s website is “The Tramways of Quito, Ecuador,” here:
 
Commercial Speed:
The stated running time for service C2, which runs the full length of the line, is 120 minutes. A colleague who has visited Quito confirms that this is the scheduled “round-trip” running time. Implied “commercial” (“schedule,” or “passenger”) speed: 16.1 km/h (= 10.0 mph).
No speed record here. Whatever else El Trole may be, it is not “rapid transit.”
 
Service Details:
El Trole operates 19 hours per day on workdays. Five overlapping services are provided over the busway. Schedules do not appear on the website, nor, according to a firsthand report, are schedules posted at stations. This is not surprising: the large number of vehicles in service implies a high service frequency.
At January 2003, the stated number of vehicles in service was 77 (out of 113) on weekdays, 50 on Saturdays and Sundays, and 45 on holidays. Vehicles are operated “one-man,” but stations have attendants.
The stated number of vehicle-km per weekday (at January 2003): 13,210 km (=8,190 mi). This implies about 172 km (106 mi) per vehicle in weekday service.
We estimate an average of (very roughly) 11 “revenue service hours per vehicle in weekday service. This in turn implies a “minimum staff requirement” of about 110 drivers (to the nearest ten) for “weekday service.” This admittedly rough estimate assumes 8-hour shifts.
 
Fare:
The stated “ordinary” flat fare for El Trole is $0.25, and the “reduced” fare is $0.12. (Ecuador replaced the sucre with the U.S. dollar as its currency in 2000.)
Tourist-information sources state that the fare for travel aboard El Trole is 50 percent less than charged by “other” buses. This helps to explain the line’s popularity: roughly 2/3 of Ecuadoran workers earn less than $30 per month.
 
Fleet:
The Quito trolleybus fleet includes 113 units, all articulated. The first group of 54 vehicles has a rated capacity of 52 seated and 114 standing passengers. The second group of 49 vehicles has a rated capacity of 42 seated and 124 standing passengers. This works out to a rated “capacity” of 166 passengers per vehicle.
Vehicle dimensions: 17.8 meters (57’ 8.6”) long and 2.5 meters (8’ 1.3”) wide. Gross floor space is therefore about 44 m2 (460 sq ft).
The “loading standard” above works out to 3.8 passengers per m2, or 9.3 passengers per meter of vehicle length. This, we believe, may be the equivalent of “maximum licensed capacity,” perhaps better described as a “load limit.”
The vehicles have three-phase asynchronous traction motors rated at 230 kW (308 HP). They also have diesel engines: 210 HP for the first group (40-gallon tank), and 308 HP in the second group (50-gallon tank).
The trolleybuses are capable of “full-speed” operation away from the overhead – and someone apparently decided this was a good idea worth “enhancing” for the second group. We believe that the vehicles were not designed as true “dual-mode” buses, as evidenced by fuel tank size: a full-size articulated diesel bus would have a tank capacity on the order of 150 gallons.
“The electrical equipment was developed with technology provided by AEG of Germany, and ‘computerized aspects’ provided by the KIEPE firm, specialists in power control systems, also of Germany. The bodies were built by the [Spanish] firm Hispano Carrocería.”
Traction current is supplied at 750V dc.
A colleague states that the “typical” speed between stops is 40 km/h (25 mph), although vehicles sometimes reach 50-55 km/h (35 mph).
 
Energy Consumption:
“The trolleybus consumes 1 gallon of diesel fuel for every 8 km operated.” This works out to about 5 miles per gallon. We believe this refers to equivalent energy consumption while operating under electric power.
 
Average Travel Distance (ATD) Per Passenger:
Estimation of this key indicator begins as follows:
260,000 passengers per weekday
/ 13,210 vehicle-km per weekday
= 20 passengers (“boardings”) per vehicle-km.
Assuming that the ATD is no less than 1 km, the number above represents the  “minimum” number of passenger-km per vehicle-km.
This implies about 20 boardings per veh-km. Unless the average travel distance (ATD) is shorter than 1 km . . . which we doubt . . . this number represents the “absolute minimum” for pass-km per veh-km.
Continuing:
1 km ATD implies 20 pass-km per veh-km.
2 km ATD implies 40 pass-km per veh-km.
3 km ATD implies 60 pass-km per veh-km.
4 km ATD implies 80 pass-km per veh-km.
We doubt that the average distance traveled by El Trole passengers is as great as 4 km. This implies an average of 80 passengers aboard each vehicle while in service, throughout the service day, which extends from 5 am to midnight.
(One of us spent several days riding public transport in Shanghai, China, in 1983. Private autos were unknown back then, several years before Shanghai received its first metro line. We think that 80 pass-km per veh-km for surface transport is unlikely –  “even” in Shanghai, “even” on the busiest transport services, “even” back in 1983 – because this is inconsistent with observations during off-peak and late-evening hours.)
A similar exercise, based on “willingness to pay” the full fare (U.S. $0.25), suggests an ATD no smaller than about 3 km. Although we believe that an ATD of 2 km is unrealistically low, we will use 2-3 km as the “likely ATD range” for El Trole. This implies roughly 500,000 – 800,000 passenger-km per weekday.
 
Weekday Traffic Density:
The weekday traffic density implied by:
--260,000 passengers per (average) weekday
--2-3 km average distance traveled
--10.1 km line length
is roughly 50,000 – 80,000 pass-km per km of route. This in turn implies an annual traffic density in the range of 13 million – 25 million pass-km per km of route.
El Trole carries a true “heavy-rail workload,” although the commercial speed it provides is 1/2 to 1/3 that which a full-scale metro might provide – and 1/4 of the commercial speed provided in some locations by Bay Area Rapid Transit, which is exceptionally fast. Relative “consumer disadvantage,” however, is not so great because access is easy (i.e. frequent surface stops) and ATD is short.
 
“Likely maximum” passenger volume:
We begin our estimation by assuming that traffic is evenly balanced by “corridor” (i.e. north and south of the centro histórico) and by direction. We acknowledge that the former is probably not “strictly true.”
The exercises above imply a “one-way, all-day” passenger volume of 65,000, which we will change to a range: 60,000 – 70,000.
The service day is 19 hours long. The “absolute minimum” hourly volume – which would require an absolutely even “temporal distribution” of travel – is therefore:
1 / 19 * 100 percent = 5 percent;
5 percent * (60,000 – 70,000) = 3,000 – 3,500.
We will round the above range to a single significant digit: 3,000 – 4,000 passengers per hour. This “should” be adjusted “downward” to some extent because not all trips begin or end in the business center (CBD). However, we do not have sufficient data on hand to permit this. We conclude that a “reasonable” minimum value for “peak-hour volume” (at the “maximum-load point”) is 3,000 pass/hr. This implies a “plausible minimum” value for the (one-way) “peaking factor” of 5 percent.
The “plausible maximum” value for (one-way) “peaking factor would be very roughly 10 percent for U.S. cities. This implies that 20 percent of “two-way, all-day” traffic moves “during the busiest hour, in the busier direction, past the maximum-load point.” Although we do not have information on transit travel patterns in Quito, we can estimate this factor to the nearest ten percent. A 20-percent “one-way” peaking factor implies a “two-way” factor of 40 percent. This can be true only if 40 percent of all weekday traffic moved during a single hour – which, to us, sounds unlikely. Because we cannot defend a “leap” up to 20 percent, we naturally retain the previous estimate. The “likely maximum” peaking factor of 10 percent implies the following:
10 percent * (60,000 – 70,000) = 6,000 – 7,000.
In other words, we estimate the “maximum one-hour passenger volume” in the range of 6,000 – 7,000 pass/hr.
Comparison with estimated “maximum service frequency” proceeds as follows: the “maximum frequency” provided over the busiest segments is reported by tourist-information sources at 2 min. This implies a “maximum service supply” of 30 vehicles per hour. This in turn implies an average of 200 – 230 passengers aboard each vehicle.
We think that an “average peak vehicle occupancy” of 200-230 pass/veh is unrealistically high – but have no information on the “maximum” service level actually operated. We believe that 60 vehicles per hour is certainly possible, and this would provide the capacity necessary to transport 6,000 – 7,000 passengers per hour. We are less certain that 10,000 pass/hr is possible. However, we believe this could be achieved if operating staff could dispatch vehicles every 30 seconds over the busiest segments during the intervals of “maximum demand” (which last roughly 15-20 minutes).
 
In Sum:
Quito’s trolleybus busway is low-capital alternative to a full-scale metro, and a “low-impact” alternative to a surface rail line (potential impacts in the centro histórico are one reason cited for rejection of surface LRT). El Trole has proven popular and practical. However, high pavement wear was reported during the early operating period. The vehicles were anticipated to last 20 years, but this may prove unrealistic because of intensive use.
El Trole, relatively slow and labor-intensive, is certainly not a system one would build in a “high-wage” operating environment to provide “auto-competitive” travel times. Also, Quiteños tolerate levels of crowding that most U.S. and Canadian transit consumers would not. Therefore, if El Trole were duplicated in the U.S. or Canada, it would certainly not carry the “same” passenger volumes as may be observed in Quito. We assert that only the most woefully ignorant planners and policymakers would ignore these issues.
 
Acknowledgements: We express sincere appreciation to Darrell Clarke for his “on-site” observations from a 2000 visit, and to Allen Morrison for descriptions from previous visits.
 
References: In addition to the links above:
“Busways: The way to go in Quito, Ecuador. Buses International, December 2002. (ht tp://w ww.busesintl.com/Dec_2002.htm; broken link at 2007.1.28.)
 
 
Addendum - Ridership:
According to an online document related to the current light metro project, ridership carried by El Trole increased as follows:
Year
Annual Ridership
Average Weekday Ridership
1996
53,422,596
164,885
2005
75,634,550
242,418