Light Rail Jacksonville

BRT -vs- LRT

Part II

BRT -vs- LRT COSTS, A REALITY CHECK

Myth:
There seems to be a widely held notion among the public – including many politicians, journalists, etc. – that rail transit systems, such as light rail transit (LRT), are weighted down with substantial heavy capital expense, while buses are more or less "free". Rail transit critics exploit this misconception by emphasizing the relatively high installation costs of new rail systems ("BILLION$$$ for rail") vs. the relatively lower costs of simply operating buses on city streets and freeways.

 "Why build expensive rail? Buses can do the same thing cheaper" is a familiar refrain in local debates over proposed new rail transit starts.

Talke a good look at this, this is from JTA and is their own image of "Cheap Bus Reality!" ...and what happens when a bus driver has a blow-out, faints, or has a heart attack, while driving on this thing?

Reality:

Bus systems incur sizable capital expenses, too, as well as rail, with typically much higher operating and maintenance (O&M) costs; often, when you add up all these costs and account for the relative life of all the infrastructure and rolling stock, plus the work performed (measured in passenger-mileage or passenger-km), you may find that rail actually gives amazing "bang for the buck".

To demonstrate this, the Light Rail Now Project team carried out an analysis of transit performance data from St. Louis Metro, comparing the total operating and maintenance (O&M) plus capital costs of both Metro's bus transit and MetroLink light rail transit (LRT) systems for the period 1996-2005, using National Transit Database Agency Profile data gathered by the Federal Transit Administration.
The table below presents total costs (capital fixed facilities and rolling stock, and O&M) for each mode over the ten-year period (millions of US dollars), and the total passenger-mileage (millions) carried by each mode over that period:

St. Louis Metro – Total Costs & Passenger Mileage, 1996-2005
 Capital Costs:
Fixed Facilities Capital Costs:
Rolling Stock...... O&M........ Cost........ Total Cost Passenger.....Mileage
Bus $64.0............ $131.3 .....$1,045.5 ......................$1,240.8..... 1,389.5
LRT $844.4 .........$124.2..... $262.2 .........................$1,231.2 .....1,047.7

[Source: Federal Transit Administration, National Transit Database, 1996-2005]

It is particularly interesting to note that, even with its heavy capital costs, when operational costs are considered, St. Louis Metro's LRT in this period exhibits total costs slightly less than the agency's bus operations. However, higher total passenger-mileage was carried on the bus system, so a more complete analysis would require taking into consideration the differing life-cycle costs for each mode (e.g., railcars last considerably longer than motor buses) by annualizing capital costs.

To obtain a total annualized cost figure for each mode, capital costs were annualized using common economic analysis (see discussion below). Annual operating costs were averaged for the 10-year period, as was annual passenger-mileage for each mode – reflecting the advantages of the longer lives of both LRT infrastructure and rolling stock.

For bus, average annual O&M costs were $104.6 million, and average passenger-mileage was 139.0 million. For LRT, average annual O&M costs were $26.2 million, and average passenger-mileage was 104.8 million.
Via this "averaging" method, with annualized capita costs, the total cost per passenger-mile for each mode was calculated as follows:

Bus - $0.88 • LRT – $0.74

This suggests that, with total capital and operational costs considered, St. Louis's "capital-intensive" LRT ends up costing approximately 16% less per passenger mile than the agency's supposedly "cheap" bus system.

This analysis was corroborated by a slightly different methodology – calculating the cost per passenger-mile for the final year, 2005, only. In this case, the annualized capital costs for each mode were added to the O&M cost for 2005, and then divided by the passenger-mileage for each mode in 2005 to obtain a total cost per passenger-mile figure for that year:
• Bus – $0.97
• LRT – $0.82
Through this method (which adjusts somewhat for more recent inflationary increases), LRT still comes out about 15% less than bus service in terms of work performed (i.e., passenger-miles carried). 

- - Followup question: Are we achieving those benefits with a 25 year, $1 billion plan?

- How will this project be funded?

These two questions are pretty well addressed by the Myth-Makers at the GAO, if you have read this far, no doubt you too will question their motives. This is a HIGHLY Fictional Summary from the GAO, in a Republican House ordered "study of BRT vs LRT" Let's put this "EXPERT OPINION" to the test shall we?

Analysis of the General Accounting Office Report "Bus Rapid Transit Shows Promise"
By Edson L. Tennyson, P.E.
GAO – Introduction: "The Federal Transit Administration (FTA) has begun to support the [bus] concept and expand awareness of new ways to design and operate high capacity Bus Rapid Transit (BRT) systems as an alternative to building Light Rail systems."
GAO – Purpose: "To examine the federal role in supporting BRT; compare the capital costs, operating costs and performance characteristics; and describe other advantages and disadvantages of BRT and LRT."
GAO – Method: "We interviewed FTA officials and industry experts to identify the advantages and disadvantages of BRT and LRT."
Comment: The "experts" are never identified but some support LRT as far more acceptable to the public and others suggest erroneous and disproven advantages of BRT.
GAO – Cost: "Two BRT projects have received funding commitments from the current New Starts program totalling about $ 831 million."
Comment: This would calculate to about $83 million per mile if the "two projects" are Boston and Pittsburgh, or $55 million per mile if they are Los Angeles and Pittsburgh, exclusive of rolling stock and garages in either case. The report's Appendix III shows that, on average, light rail projects in Denver, Sacramento, Salt Lake City, San Jose, and Saint Louis actually cost $24 million per mile at current dollar values including rolling stock and shops. However, GAO increases its LRT cost to an average of $ 34.8 million per mile by including subways (included in a few systems); yet the costly bus (BRT) subway in Seattle (above right) was "inadvertently" omitted from the study. Reserved street lanes were included in the BRT average to bring the cost down but these are in no way "rapid transit".
GAO – Public Preference: "Transit officials repeatedly noted that buses have a poor public image...."
Comment: The report also notes that "Los Angeles and Houston originally built part of their [BRT] systems as exclusive busways and later converted them to HOV [High Occupancy Vehicle – automobile and bus] facilities." (Northern Virginia could be added to this list.) Politicians perceived too little use of the costly busways, but with HOV in Los Angeles, buses were delayed and lost much patronage until HOV rules were made more restrictive. In Northern Virginia, HOV rules are occasionally suspended and bus movement stops or is severely delayed, making second trips of the bus impossible.
GAO – Misinformation: "The reason for the relatively few BRT projects being ready for funding .... includes the newness of BRT."
Correction: BRT is hardly "new". The first known BRT operation was in 1938 in Newark, New Jersey's Cedar Street subway. Service was discontinued after gasoline rationing ended. The second known BRT was the Ardmore Route 103 of Philadelphia Suburban Transportation Co., converted in 1967 from LRT Costs of operation increased and 15 percent of the passengers were lost. Northern Virginia opened their BRT in 1970 and has lost riders ever since the energy crisis of 1980, now down about 67 percent. Pittsburgh's first BRT was in 1977, before any new LRT starts in the USA. Planners promised 32,000 weekday passengers but the service now averages 14,500.
GAO - Policy: "FTA promotes the BRT concept with the slogan 'think rail, use buses'."
Correction: The implication that buses provide equivalent attributes or advantages to those of rail transit, and at lower cost, is deceptive. The GAO report's Figure 6 shows average BRT (construction) cost at $13.49 million per mile, but this involves the simple marking of street lanes for 12 mile an hour service, not really BRT. See the discussion of 'Cost' above, which notes the cost of recent busways at $55 or $83 million per mile. Conversely, GAO's purported LRT "average" of $34.8 million per mile includes some subway or tunnel construction which was omitted from the BRT average cost.
GAO – Misinformation: "In Pittsburgh they needed to construct only five [5] miles of busway to achieve the goal of rapid transit to the airport because the buses could exit the busway and use existing highways." The report adds that an alternative for light rail would have been longer, "cost two to three times as much to construct and equip for operation without attracting any significant added patronage."
Comment: Pittsburgh's so-called "Airport Busway" (see photo at right) goes nowhere near the Airport (which has a freeway access) so has been renamed the West Busway. The cost estimate of $320 million was for eight miles, but the bids came in at $515 million. To fit the funding, the busway was cut to five miles, eliminating the greatest congestion relief nearest to downtown.
The Buffalo subway was used for LRT costing without a detailed study, but the Pittsburgh project was on abandoned railroad right-of -way. A current LRT project in Pittsburgh is costing about $325 million for five miles of very costly hillside construction, but with subway access to downtown. The West BRT was estimated to attract 50,000 weekday passengers but so far has attracted about 6,000. An independent estimate for rail operation forecast 20,000 weekday passengers.
GAO – False Statement: "We also attempted to determine operating cost per passenger-mile as a measure of comparison, however we could not obtain sufficient data for such an analysis".
Comment: FTA has required passenger-mile data at least since 1982. GAO reported the cost per hour was $84.72 on its selected bus systems and $161.48 on LRT systems in the same cities. GAO also reported buses may carry fifty passengers and LRT 110. Actual capacity is greater but these numbers are realistic (6.8 square feet per passenger average).
Thus the bus cost per passenger for an hour ride is $1.69 and the LRT cost is $1.47 for the same trip. If the average speed is 15 miles per hour, and passengers ride the full route length, the cost per passenger-mile would be 11.3 cents by bus and 9.8 cents by rail. However, actual passengers seldom ride the full length of the route, so the passenger-miles must be cut in half, raising the bus cost to 22.6 cents and LRT to 19.6 cents per passenger-mile.
The actual cost of express buses in Dallas was 46 cents per passenger-mile, the only city where the data were segregated out. In San Diego and Saint Louis (and Calgary in Canada) LRT cost was 18.5 cents per passenger-mile in 1998. Did GAO avoid these well-performing systems on purpose?
Dallas express buses are contracted out to lower-labor-cost operators, so their costs cannot fairly be compared to LRT there. Articulated buses could not be compared for lack of data, but an operations research study of actual operations in Los Angeles found that articulated buses cost more per passenger than standard buses because of slower operation, and higher maintenance and safety costs.
Erroneous Data: GAO's Figure 7 is replete with errors. Los Angeles's cost per bus hour is not $56 but $93.72, according to FTA in 1998. LRT in L.A. is not $434 per hour, but $253.94. This invalidates the GAO findings.
The GAO report's Figure 8 also contains several errors involving the cost per mile of operation. The Dallas GAO cost of $1.74 for buses was published by FTA as $7.24, or $3.48 for the express buses with higher speed and lower labor cost. The cost per mile varies inversely with speed. Denver's bus cost per mile was said by GAO to be $2.24, but FTA reports $5.10. Denver has many long, fast routes. GAO reported Los Angeles at $3.45 per mile, but FTA's published figure was $8.01.
LRT costs related by the GAO report were generally correct. It is possible staffers gave GAO their estimate of "avoidable' or incremental costs whereas FTA reports the full cost. GAO Figure 9 is similarly erroneous. Dallas bus trips do not cost 31 cents as stated, but $3.19 local and $3.68 express. Perhaps a decimal point was misplaced. Denver cost $2.38, not $1.21 as reported. Los Angeles was $1.72, not $1.06. Here, too, we may be seeing avoidable costs, not full costs.
Ridership: GAO found that "the largest ridership on BRT and LRT is quite similar. Busways averaged about 15,000 riders per day" but "LRT system ridership ... averaged about 29,000 per [week]day.' There is no way 15,000 is "quite similar" to 29,000. In the Los Angeles CBD at rush hour, ridership on the street averages 32,500 per day, but it is a fact that the "weak" LRT Green Line averages 33,000, nowhere near the CBD, and the Blue LRT Line to the CBD averages well over 63,000. That is why "transit officials" told GAO that passengers prefer rail.
Speed: The GAO report's Figure 10 shows a 56-mile-an-hour BRT speed, but that must be without any stops for passengers. LRT is listed at 16 miles per hour which is realistic for the slower lines with the most stops and 36,727 weekday passengers. However, that speed includes the turn-around time at each terminus; the actual schedule speed is about 20 mph. The Dallas express bus system averaged 17 miles per hour in 1998 with many routes in all directions, but carried only 31,464 weekday passengers. There are only two LRT routes in Dallas, but they attract more passengers.
Denver received the most distorted portrayal by the GAO study, with LRT being reported at only 11 miles per hour. This was for the downtown segment of the route. The full route speed is 23 miles per hour. The bus speed of 35 miles per hour applies to the isolated freeway portion of the route.
The entire Los Angeles Green LRT Line also averages 35 miles per hour, with stops, for 33,000 weekday passengers. Rail cars have faster acceleration than bus, particularly on grades, where bus drivers sometimes turn off the air-conditioning to avoid very slow operation. This is not a problem with electric rail.
In Northern Virginia, the Shirley HOV-Busway is non-stop, but when off-line stops are added, the speed falls to 22 miles per hour, about the same as the schedule speed of many LRT lines. However, the Shirley service attracted only 12,868 weekday passengers in 1992, with losses since then. LRT averages about the same on suburban lines, but with far more patronage. Most BRT lines have limited or no weekend service for lack of demand.
Flexibility: GAO correctly reports that BRT excels in this attribute, but does not explain that passengers do not respond well to flexibility as noted above. The fixed right-of-way is an advantage that experience has clearly proven.
Gradual Phase-In of BRT: This purported "advantage" is claimed by GAO, but this is in reality a great disadvantage for passengers who are the purpose of the service. Benefits and savings are ignored.
Negative Busway Attributes: GAO admits that buses are not popular with the public, "particularly when compared to rail service." GAO seems to assume that buses can be made popular by calling them BRT. GAO ignores real-time experience.
For example, when the local trolley service (see photo at right) replaced San Diego Express Bus Route 90 on freeway 94, ridership increased from 3,000 per weekday to 12,000 east of Euclid Avenue. Protests about the switch to local service were soon forgotten. When Denver replaced express bus service on Santa Fe Drive (US 85) with all-stop LRT service, ridership more than doubled and weekend service became economical.
Many LRT passengers avoid transferring by driving to park-and-ride facilities. Buses do not attract such riders at the same rate, but they do attract some of them. In Cleveland, Detroit, and Saint Louis, extensive freeway express bus service has been decimated, eliminated on many routes, but LRT has done very well, and at much lower operating cost per passenger-mile, except in Detroit where there is no LRT. Detroit's Pdpwntown People Mover (an automated guideway transit system) is very high cost, with low ridership, despite its exclusive guideway.
Light Rail Advantages: These are largely passed over by GAO. While conceding the economic development advantages, no mention is made of travel safety whereby bus service experiences 80 percent more injuries per million passenger-miles. The lack of LRT air pollution is not mentioned, nor are labor savings with train operation.
The GAO report does not explain why the West Busway in Pittsburgh, the Harbor Freeway Transitway in Los Angeles, and the Seattle bus tunnel are omitted from the report's data, other than to say the West Busway is too new. That did not stop the authors from touting the advantages of unbuilt systems. These three facilities cost $1.2 billion to build to cover 18 miles, $67 million per mile with only 34,000 total weekday passengers on all three. The L.A. and Pittsburgh projects have attracted only 9,000 total weekday passengers combined, but 113,000 were promised. Why was this not reported?

WOW! Now this WAS a walkable city!  Yes, "We have met the enemy and they are us!" "The Most Beautiful Streetcar line In The World," this once was  North  Main St., Jacksonville!