Caen: Guided BRT out, real LRT tramway in by 2019

Rendition of Caen's proposed LRT tramway that will replace problematic guided-BRT system. Graphic: Caen municipality.

Rendition of Caen’s proposed LRT tramway that will replace problematic guided-BRT system. Graphic: Caen municipality.

When the “tram on tyres” or “rubber-tired tramway” technology first emerged in the early 2000s, it was positioned as part of the new Bus Rapid Transit (BRT) concept attracting interest at the time. The argument went that “BRT” was “just like light rail, but cheaper”, and the “rubber-tired tramway” was intended to demonstrate that a “tram” constructed with automotive/bus technology could be “guided” just as a light rail transit (LRT) tramway was guided by its track rails, and able to operate extra-long, multi-articulated buses smoothly and reliably just like the tramcars on LRT railways.

A number of cities have experimented with or adopted the technology, particularly in France, where cities like Nancy, Clermont-Ferrand, and Caen made the “tram on tyres” the centerpiece of their transit systems. Now, plagued by reliability and performance problems, Caen is clearly fed up with it, and has launched a project to convert to a standard LRT tramway — running on bona fide tracks — by 2019.

In France, the designation TVR, Transport sur Voie Réservée, roughly translated as “transport on reserved way”, is used to refer to these rubber-tired guided-bus systems. In English, they’re often referred to as GLT, for Guided Light Transport. As explained in a Wikipedia article, “GLT vehicles bear a strong resemblance to trams, but are actually buses capable of following a single guidance rail or even operating without any surface guidance system.”

Opened in 2002, Caen’s guided-bus system eventually stretched to 15.7 km (9.8 miles), using longer-than-usual articulated buses guided by a flanged wheel running on a center guiderail in the middle of the paveway. While the buses have diesel motors (and steering wheels, so they can be driven to their garage at night), their ordinary propulsion is electric power, via an overhead contact system (OCS) and LRT-like pantographs, with the guiderail also serving as the electrical return circuit. (The dual rails of standard LRT serve this same purpose.)

Caen guided BRT ("rubber-tired tramway") system, now scheduled for replacement by LRT. Photo:

Caen guided BRT (“rubber-tired tramway”) system, now scheduled for replacement by LRT. Photo:

However, reliability problems with the technology (especially derailments of the guidewheels) reportedly have persuaded Caen’s political leadership and transit management to ditch the guided-bus system. In the new LRT tramway plan (see graphic simulation at top of post), 16.8 km (10.4 miles) of LRT routes will replace (and slightly extend) the guided-bus routes, and tracks will replace the paveways (or be embedded in some sections of pavement). A fleet of 23 trams is projected to replace the BRT buses, with a total project cost estimated at €247 million (currently about $269 million, or about $26 million per mile). Project completion is aimed for 2019.

From its early years, the usefulness of the system, as a substitute for standard LRT, baffled transit advocates and professionals. As John Carlson, one advocate posting to the Eurotrams list in 2004, commented

I found the system at Caen and also the one at Nancy to be a solution in search of [a] problem. While there must be some economies from installing just a guide rail instead of double-railed load bearing track looking at the system in [situ] I would have to ask if the guide rail is needed at all.

The vehicles are long and do turn some sharp corners but I’m still not sure if they would be beyond a competent driver and a well-constructed articulated bus operating without a guide rail.

As time went on, other problems, such as pavement wear, began to emerge. Graeme Bennett, a transit advocate in Melbourne, posted observations about the Caen system in the summer of 2005:

A friend and I recently visited Caen and were shocked, stunned, and amazed as we watched and rode these weird vehicles.

We found they were speedy, but fairly noisy, and seemed to do the job well, although they rode more like a trolleybus rather that a tram, in particular with a lot of vertical perambulations and rear end whip as they rounded corners at speed!!

One point that was obvious is the fact that because the vehicles follow exactly the same part of the road without any deviation for cut in or out, … the road surface in some areas is becoming badly damaged particularly at some of the stops where it was noted repairs have had to be made.

Even the smallest pothole will deteriorate rapidly and every tyre on every bus will hit that spot in exactly at the same place every ten minutes or so.

Bennett also observed what seemed to be an emerging problem in keeping the guidewheels in contact with the center guiderail, reporting that “We noted several “Rerailers” around the system to direct the guides onto the track.”

By 2009, serious problems with derailments were being experienced. At the end of May that year UK transit advocate Simon P. Smiler reported that, days earlier, “there was another derailment in Caen, and now it seems that their TVR rubber tyred ‘trams’ are only providing a part time service.”

Smiler wondered “Will this result in the ultimate death of the TVR as a mode of transport? Caen was looking to getting more TVR’s to expand its system — so what will it do now?”

Caen’s experience re-opens anew some of the considerations we originally raised 15 years ago in our article prompted by the very similar new guided-bus system in Nancy (also plagued with guidance reliability problems): «“Misguided Bus”? Nancy’s BRT Debacle Exposes Pitfalls of “Half-Price Tramway”». Asking “Does the ‘guided bus’ really have a purpose in life?” our article pointed out that

They basically will have a system of elongated trolleybuses camouflaged as “trams”, with lots of gadgetry to keep the buses on course. They will have a central slot to deal with in the middle of the paveway (tending to collect rain, mud, etc.). And they will be persistently trying to solve lots of operational challenges over the next months and years to prove the whole thing works. Thus one can safely predict that Nancy will be expending a lot of its planning and administrative energy trying to solve the challenges of making a trolleybus system mimic the performance of an LRT system.

There’s a recurring question: Why bother at all with the guide rail in the slot? it is dubious whether such an arrangement will permit higher vehicle speeds, although Nancy designers seem to think their bus will run a bit faster in a narrow right-of-way if it’s guided in this fashion. One is tempted to suspect that the extra-long, multi-articulated bus benefits from having its axles guided by such a mechanism, possibly minimizing any misalignment of the rear section while in the guideway (which might explain why the vehicle tends to “fishtail” when free-running).

And beyond the question of whether it’s worthwhile trying to imbue a bus with LRT characteristics, there’s another issue as well. Once a transit agency or government entity buys into an entire, specific “guided-bus” technology, its planners and decisionmakers commit to a specialized guideway and technical infrastructure using one form or another of specially designed curbs, below-pavement conduits, special travel lane markings, etc. That might happen after the initial order of vehicles, where competition is alive and well, and the initial bidding environment may be fairly competitive among a number of vendors.

However, the agency then has a stock of specialized buses with a 12 or 15-year life expectancy and capital costs sunk into building a specialized guideway which may work properly with only one manufacturer’s product. When the agency proceeds to expand the fleet or must find replacement buses, it may well find itself “trapped” with only one manufacturer/bidder. Is any vendor going to assure transit planners that its proprietary technology will become an industry standard in the next dozen years?

In contrast, imagine instead that the transit agency set down a few miles of steel rails with 1435 mm (standard) track gauge with readily available, dependable track switches, and mature signalling technology. The agency buys a couple of dozen light rail vehicles which have a lifespan of 30 to 50 years with trainlined controls so that one operator can control two to four cars. When it’s necessary to expand that system or replace the vehicles, the agency will find at least half a dozen suppliers lined up who can make cars which will work fine with the previous generation. Productivity is better, competition is alive and well, and the technology is mature.

Certainly, in view of recent experience, those comments seem as relevant today as they were a decade and a half ago. ■

Latest FTA data: Light rail trumps “BRT” in key performance measures

Left: Portland MAX LRT. (Photo: L. Henry). Right: Cleveland Healthline "BRT". (Photo: GCRTA).

Left: Portland MAX LRT. (Photo: L. Henry). Right: Cleveland Healthline “BRT”. (Photo: GCRTA).

Until recently, industrywide comparisons of performance between light rail transit (LRT) and the specific bus service mode of “bus rapid transit” (“BRT”), relying on reporting information in the National Transit Database (NTD) of the Federal Transit Administration, have been impossible because “BRT” data were not separately reported but instead were merely jumbled into the large general category of Bus. However, that has recently changed.

A number of transit agencies are now reporting “BRT” performance data within a separate category, with a total of seven agencies specifying their “BRT” data in the 2013 NTD report (the most recent so far). Thus it’s now possible to perform an analysis of LRT vs. “BRT” data to produce a preliminary evaluation of comparative performance of the two modes. (Because of the wide disparity in infrastructure and operational conditions applied to “BRT”, Light Rail Now continues to refer to this diversely and hazily defined modal designation within quotation marks.)

A comparative analysis of these “BRT” data and available data for recent-era new LRT systems (defined as post-1970, roughly following the introduction of the LRT concept in the North American transit industry) indicates that new LRT systems continue to excel in the two most critical performance areas of ridership and operating and maintenance (O&M) cost per passenger-mile. New recent-era LRT systems included in this analysis are those in the following cities/metro areas: San Diego, Buffalo, Portland, San Jose, Sacramento, Baltimore, Denver, St. Louis, Los Angeles, Dallas, Salt Lake City, Minneapolis, Houston, Phoenix, Charlotte, Seattle, and Norfolk. However, New Jersey Transit’s Hudson-Bergen LRT (HBLRT) system, launched in 2000, could not be included in this analysis of totally new systems, because the data for HBLRT is combined with that of Newark’s legacy subway-surface LRT system in the agency’s NTD report.

“BRT” systems with NTD data available include those in the following cities/metro areas: Cleveland, Eugene, Los Angeles, New York City, Kansas City, Las Vegas, and Orlando. Note that a number of important new “BRT” operations, particularly those in Pittsburgh, Miami, Seattle, Honolulu, Charlotte, Boston, and Ft. Collins, are not included because their specific data are not reported to the NTD.

For more than two decades, proponents of “BRT” have pursued a virtual war against LRT with the mantra “just like light rail, but cheaper” — claiming that an array of rebranded and heavily promoted limited-stop bus services, deployed service applications similar to those of LRT, could offer all the benefits at far lower cost. Such claims can now be tested by comparing very similar relatively new installations of both systems. Derived from a comparative analysis of this data population, critical performance indicators are presented and discussed in the sections below.

Ridership — Certainly, average annual ridership is one of the most important indicators of a transit operation’s performance. As Exhibit 1 indicates (below), in this comparison of similar installations LRT services attract approximately three times the average annual ridership of “BRT”. However, it should be noted that the majority of LRT systems have been operational longer than the “BRT” systems.

Exhibit 1. Ridership comparison.

Exhibit 1. Ridership comparison.

Another important performance indicator is ridership per route-mile (or route-kilometer). This could be calculated from “Fixed Guideway Directional Route-Miles” in the NTD. Unfortunately, while these were available for LRT, none of the “BRT” systems presented this data in the 2013 report. Perhaps this data will be reported in future NTD reports.

O&M cost per rider-trip — In this important performance indicator, the “BRT” systems in this study averaged significantly better — 38% lower — than LRT, as shown in Exhibit 2. However, a drawback of this metric is that it fails to account for differences in average trip length, as discussed in the other performance indicators further below.

Exhibit 2. Comparison of O&M cost per rider-trip.

Exhibit 2. Comparison of O&M cost per rider-trip.

Another problem with this metric: While each agency’s LRT is a “closed” system (including virtually all costs, from platform operations to vehicle and way maintenance) with operational expenses compartmentalized and accounted for, “BRT” way maintenance accounting varies from agency to agency — sometimes funded by the transit agency, sometimes by the city or county in their public works budgets. Other “BRT” expenses, such as vehicle maintenance, may be blended with systemwide bus expenses. Likewise, while LRT security operations are almost always controlled and financially allocated to the LRT budget, for “BRT” this item may be hidden in systemwide costs. All told, there is really no consistency in how some “BRT” expenses are tallied and reported, thus affecting comparability to LRT costs.

Average trip length — Differences among modes may have different influences on passenger behavior and preferences, resulting in characteristically different average passenger trip lengths. This may also affect cost per passenger-mile. For example, the average O&M cost per trip of regional passenger rail operations is often compared disadvantageously with that of urban modes, including bus operations. However, the units cost per passenger-mile may be lower as longer trip lengths are factored in.

As illustrated in Exhibit 3, analysis of the 2013 ATD data indicates that comparable LRT systems attract passenger trip lengths almost exactly twice as long as the “BRT” systems in this study.

Exhibit 3. Comparison of average passenger trip length.

Exhibit 3. Comparison of average passenger trip length.

O&M cost per passenger-mile — This unit-cost metric is by far the most important indicator for assessing financial performance, since it measures the actual work being performed — the actual transportation of passengers — rather than cost based on merely the number of “bodies” boarding the average transit vehicle. As shown in Exhibit 4, The LRT systems in this study averaged an O&M cost per passenger-mile approximately 17% lower than the “BRT” systems reported.

Exhibit 4. Comparison of O&M cost per passenger-mile.

Exhibit 4. Comparison of O&M cost per passenger-mile.

The bottom line: In critical metrics of transportation activity, LRT continues to demonstrate major advantages.

NOTE: Since original publication, this post has been revised with a modification to the graph of cost per passenger-mile data (Exhibit 4). The original scale ($0.48 to $0.66) has been changed to $0.00 to $0.70 to reflect a minimum zero-value consistent with the other graphs. Also, in the discussion of O&M cost per rider-trip, a section has been added explaining the difficulty in accounting for some “BRT” expenses. Rev. 2015/07/02.

Bus operations as precursors of light rail transit

Seattle — Link LRT trains (left) and BRT buses share the Downtown Seattle Transit Tunnel, originally installed as a busway.

Seattle — Link LRT trains (left) and BRT buses share the Downtown Seattle Transit Tunnel, originally installed as a busway.

by Lyndon Henry

High-quality bus services – often characterized as “Bus Rapid Transit” (BRT) – are frequently portrayed as possible precursors of electric light rail transit (LRT) systems. But can BRT or “BRT-like” bus operations effectively fulfill such a role?

This question was examined in a paper that I and my colleague Dave Dobbs (executive director of the Texas Association for Public Transportation, and publisher of the Light Rail Now website) presented to the Joint International Light Rail Conference co-sponsored by the U.S. Transportation Research Board and the American Public Transportation Association in April 2009 in Los Angeles.

The formal paper, Bus Rapid Transit as a Precursor of Light Rail Transit? is available online, starting at p. 137 of the full conference proceedings:

Our study includes general research as well as an examination of several specific case studies, drawn from both actual operations and planned operations. Our analysis identifies factors that may optimize the capability of of these kinds of bus operations to function more effectively as precursors of LRT systems, but emphasizes that “initial system design, to permit a transition, is critical, and major challenges and drawbacks must be addressed and overcome.”

Our conclusions spell out some details:

This analysis has identified certain factors that may optimize the capability of Bus Rapid Transit to function effectively as a precursor of a light rail transit system. However, it is important for planners to keep in mind that initial system design, to permit a transition, is critical, and major challenges and drawbacks must be addressed and overcome. A major consideration is that the BRT facilities should not represent an obstacle to the subsequent LRT project. As noted, BRT-specific infrastructure (including stations) should ideally be designed to be very low in cost so the sunk cost for BRT is not an impediment to eventual conversion to LRT.

The examples of actual or prospective BRT-to-LRT conversion in both Seattle and Ottawa (and in fact Guadalajara as well) involve some degree of transit service shutdown or disruption on the BRT facility during the conversion process in these types of “high-end”, exclusive facilities. In contrast, a “lower-end” express-bus type of BRT service can probably more readily continue a parallel service on adjacent highway or arterial lanes (if they are available) during the conversion period – although generally without stations and intermediate interchange of transferring passengers (an essential characteristic of LRT which planners should seek in BRT if the BRT service is intended to offer really the same kind of service as LRT).

In addition, the staging and logistics of conversion must be considered, particularly to avoid or minimize disruption of the existing BRT-type service while the LRT installation project is under way. In this regard, alignments in or alongside existing arterials provide at least some opportunity for maintaining a parallel BRT or bus-substitute service; on the other hand, alignments that have appropriated railway ROW for BRT (such as the Ottawa Transitway) make it virtually impossible to maintain a true parallel bus service – thus representing a serious obstacle facing conversion to LRT.

On the whole, the case studies cited suggest that actual experience is still inconclusive as to full cost-effectiveness of some forms of BRT service functioning as precursors to LRT and other type of rail transit. However, several examples approaching implementation in the near future appear to show show promise. As these planned BRT-to-LRT conversions become operational, an updated assessment should be carried out.

For additional information, plus useful graphics, check out the PowerPoint presentation shown at the conference:

Bus Rapid Transit as a Precursor of Light Rail Transit?