Estimating PRT Costs
Estimating the cost of a PRT system is relatively easy if one is willing to settle for a general "ball park" number within 20% of actual costs. Two primary methods for estimating costs suggest themselves: 1) review the costs for whatever PRT prototypes have been built, and 2) find projects similar to PRT in terms of size, weight, and complexity. For purposes of estimating the costs of PRT in transit projections, the following leads to a conclusion that PRT costs between $10M/mile and $15M/mile.
The Advanced Transit Association (ATRA) offers a good starting point with their reports on "Personal Automated Transportation: Status and Potential of Personal Rapid Transit". A table on page 89 of their Technology Evaluation (dated January 2003) presents cost estimates from PRT vendors. Of the 14 systems reporting cost estimates, four have actually built a prototype: Cybertran, Frog (CyberCab), Taxi 2000, and ULTra. Most of the other estimates are lower than these, probably because of 1) missing project costs (site prep, utility relocation, foundations, landscaping, etc.), or 2) the assumption of some level of (as yet unattained) mass production.
For those four, guideway costs range from $2.6M/mile for Taxi 2000 to $5.0M/mile for Cybertran.
Assuming a saturation level of one vehicle every 100 feet (or 50 per mile of guideway), vehicle costs range from $2.1M/mile for Taxi 2000 ($42,000/cab) to $8.0M for Cybertran. [Note that Cybertran vehicles carry 6-20 passengers while Taxi 2000 and ULTra vehicles hold a maximum of 4 passengers. The smaller vehicles and superstructure likely account for the lower costs of both Taxi 2000 and ULTra.]
As indicated in the "Additional component costs" on page 90, station costs can range from $100K to $500K per station. Stations would be ADA compliant and usually include elevators. Planning for two stations per mile would add, at most, $1M/mile to guideway costs.
Costs for land acquisition are excluded because public right-of-way (primarily roadways) will be used. Due to the small footprint of each supporting post (approximately 7 square feet every 60-90 feet) and the flexibility of routing, utility relocations will be minimal and relatively insignificant. Other costs might include modifying buildings, providing parking at stations, and enhancing aesthetic features. Adding an extra 20% for planning, engineering, environmental studies, and project management seems reasonable (based on Figure 8 entitled "Component Cost Distribution" in the document at http://faculty.washington.edu/jbs/itrans/yoder.pdf). So, the minimum cost would be $2.6M/mile(guideway) + $2.1M/mile(cabs) + $1.0M/mile(stations) + 20%(overhead) = $6.8M/mile.
In summary, full costing for these four systems can be expected to range from $6.8M/mile to $16.8M/mile. Bi-directional costs, which are useful for comparison to corridor-type transit systems, would be $13.6M to $33.6M/mile. Note that the high end of the range is set by Cybertran which, due to size and design, costs more than true PRT systems.
[ATRA's report, "Personal Automated Transportation: Status and Potential of Personal Rapid Transit can be found at www.advancedtransit.org/advanced-transit/studies/]
Another estimate comes from a recent article in the Feb. 2004 "Mechanical Engineering Design". The ULTra system is estimated to cost $8-12M/mile (one way). Here's an excerpt from page 33:
In Cardiff, a half-mile test track for ULTra has been up and running since 2001. Lowson and his team have taken to heart the lessons from Raytheon's failure. "Raytheon didn't pay enough attention to the infrastructure," Lowson said. "It got seriously oversize and overweight and therefore over cost. We've spent a lot of time looking at optimizing the infrastructure." For example, the depth of the steel guideway is only 18 inches, which saves on material both in the guideway itself and in the structures that support it.
"The span-width ratio exceeds the normal rule of thumb, but we can exceed them because our loads are very different," Lowson said. "It's not a footbridge, and it wouldn't work as a footbridge because it's too thin. But it works as a PRT guideway."
And unlike Raytheon, which developed all the technology for its system from scratch, Lowson's team concentrated on finding off-the-shelf parts to build ULTra. Lowson's group built the vehicles from automotive industry parts, which enabled them to piggyback on decades of research. Consequently, Lowson says that ULTra has greater reliability and lower costs than other transit systems.
As a result, Lowson says he can deliver ULTra at a cost of between $8 million and $12 million a mile, far cheaper than conventional light rail systems. And in comparison to relatively inexpensive solutions such as diesel buses, ULTra can theoretically carry far more people per hour. Based on rider studies conducted in Cardiff and on estimates of the capacity of the line, the system could deliver about 2,000 people each hour among a dozen stations.
[The Mechanical Engineering Design article entitled "Ready when you are?" can be found at http://www.memagazine.org/medes04/readywhn/readywhn.html]
Smaller vehicles allow a light- weight guideway with total system costs estimated at under $15 million per mile.
SmartSkyways offers their own "Estimate of Costs for Skyways Infrastructure" at http://www.smartskyways.com/Technology/development_costs.htm.
"Capital Costs and Ridership Estimates of Personal Rapid Transit"
by Supin L. Yoder, Sidney E. Weseman, and John DeLaurentiis (January, 2000)This report examines the personal rapid transit (PRT) system project in Rosemont, Illinois. A method was developed to compare PRT system components (versus the entire system) with the components of existing automated-guideway transit (AGT) and automated people mover (APM) systems. Components examined included (1) guideways; (2) stations; (3) maintenance and control facilities; (4) power and utility systems; (5) vehicles; (6) command, control, and communications systems; and (7) engineering and project management. In their examination of 17 AGT systems, they created Figure 8, Component Cost Distribution which shows that the average cost distribtuions were guideways (27%), stations (10%), maintenance and control facilities (6%), power and utility systems (7%), vehicles (18%), command, control, and communications systems (14%), and engineering and project management (18%). Their conclusion: "Raytheon’s PRT capital cost estimate is within the range of $27-$37 million/lane-mile of PRT-comparable AGT system cost."
Both this range and the report's broader range of $15M/mile to $50M/mile are much higher than the other estimates on this Web page. Three aspects of the report call into question its high estimates:
Although the report’s lowest estimate for PRT is $15M/mile, the Clarian APM cost only $14.2M/mile. PRT, being much smaller and lighter than the Clarian APM, would likely cost less. (See Clarian photos at http://www.sibelle.info/peoplemover/carpics711.htm)
Table 2 of the report, “Major Factors Impacting Component Capital Costs”, shows that all PRT components lie on the “LOWER COST” side with these exceptions: linear induction motors (guideway), and ADA compliant (vehicle), freestanding (station), ADA/Elevator & Escalator (station), Fare collection (station), and CCTV, Intercom, PA (station). PRT, with so many lower cost attributes, would likely lie at the lowest end of the range, not the middle.
Smaller and lighter has a ripple effect. As some or most parts of a guideway or vehicle become smaller and lighter, other parts can also be scaled down. Smaller vehicles lead to smaller guideways and stations. This synergy toward small, light, and efficient – and its resulting lowering of capital costs - isn’t addressed in the report.
[The article can be found at http://faculty.washington.edu/jbs/itrans/yoder.htm]
[The Figures can be found at http://faculty.washington.edu/jbs/itrans/yoder.pdf]
According to the October 2002 issue of TeMPO published by the Metropolitan Planning Organization for the Indianapolis, IN, region:
The Clarian People Mover is America's first privately owned transit system to operate over city streets. Each vehicle can seat 8 and stand another 19. A train-set includes 3 vehicles for a total of 81 people. Each fully automated trainset will travel at up to 30 mph on rubber wheels riding on pre-stressed concrete guideways. Empty weight of each trainset is 45,000 pounds. [Note that each Skyweb Express cab weighs only 1000 pounds.] Even though a 1700" section of an 18" gas main had to be moved laterally 5 feet, the overall cost of the system was only $40M. The Clarian People Mover is 14,800 feet long (7,400-foot long two-way guideway), or about $14.2m per one-way mile. Operating costs are estimated at $900K per year (2.25% of construction cost). The project broke ground in May of 2001, and the superstructure was completely erected by mid-2002, right on schedule. The system was commissioned in the Spring of 2003. System capacity is 1800 riders per hour with an expected 500,000 riders per year.
Although most APM systems run $20-30M/mile (world-wide estimate of systems in development or recently completed), this one cost only $14.2M/mile. A large part of the difference can be attributed to the smaller, lighter design. However, the fact that the Clarian system is privately built (and therefore more financially accountable) may explain a significant portion of the cost reduction.
[View other photos at http://www.sibelle.info/peoplemover/carpics711.htm.]
Another relevant recent costing study was BART's Group Rapid Transit Investigative Study. Although GRT differs from PRT, an independent GRT costing provides a reasonable upper limit for PRT costing. In the BART study, one of the subs on the SFO people mover came up with costs for a 5-mile system for Cybertran's GRT system. The guideway structure is rated to sit on the Oakland Airport fill and withstand a big earthquake. Although the final report has not been finished, the consultant's version of the costing should be available to MTC from BART. Note that Cybertran vehicles carry 6-20 passengers while Taxi 2000 and ULTra vehicles hold a maximum of 4 passengers.
Another common technology that in many ways resembles PRT is amusement and "roller coaster" rides. Here's an excerpt from an e-mail sent by Dick Gronning, a PRT supporter:
I called S&S Mfg. that does roller coasters. Their smaller track can be had for $3.2 million a mile, which is about what Taxi 2000 says its track should be. However, S&S produced a track in Japan that looks very similar to the Taxi 2000 track and cost $7.1 million a mile [see photo]. This track regularly takes twenty riders to over 110 mph with forces five times normal gravity. This is a bit more than what Taxi 2000 needs.
Then I talked to Met-Con Construction in Faribault about the price of stations. The estimator engineer said that a three car station would be $150 to $200 thousand with stairs and elevator. He said that a five car station would be between $400 and $500 thousand per station. Taxi 2000 uses figures of $400 to $600 thousand for stations.
A harder estimate is the cars. A small Ford Focus, or Chevy Malibu can be purchased at about $10 thousand, but these cars are manufactured at a rate of 20 thousand a month. They are also much more complicated and have far more moving parts than a Taxi 2000 car. So, what could a car like Taxi 2000 with only sixteen moving parts and only produced at a rate of two thousand a month really cost? Taxi 2000 says that if they could get someone to manufacture 2000 cars a month, the cost would be about $25 thousand, but if they only produced 1000 cars, the cost would go up to $30 thousand per car.
Councilmember Dean Zimmerman has proposed a system for Minneapolis that has 31 miles of track, 68 stations and probably can use about 2000 cars. I calculated the cost using the highest figures from the estimates above and got $320.8 million, or $10.3 million a mile.
One could also examine the costs of systems that WGH has built at a UK mall, a UK entertainment complex, and museums in Helsinki and Abu Dhabi.
Transportation vehicles vary from light to heavy. Generally speaking, cost increases with weight. (Thanks to Larry Fabian for suggesting this and providing background.)
Bicycle
20-40 lbs.
Electric Bike and Scooters
30-150 lbs.
Motorbike, Scooters, Motorcycles
200 - 400 lbs.
Personal Rapid Transit (PRT)
800 - 1500 lbs.
Automobiles (cars, taxis, mini-vans, SUVs)
2000 - 8000 lbs.
Buses
xx,xxx lbs.
Group Rapid Transit (GRT) and Automated People Movers (APM)
9000 - 20,000 lbs.
Light Rail Transit (LRT)
80,000 lbs.
Heavy rail (Caltrain, ACE, BART)
xx,000 lbs.
Based on the above figures, one can reliably predict that fully costed PRT will run about $12.5M per mile. Utilizing a "fudge factor" of plus or minus 20% gives us a cost range of $10M/mile to $15M/mile. Although somewhat higher than most PRT estimates, this range provides an added degree of confidence until an actual system is deployed. For purposes of estimating the costs of PRT in transit projections, transit professionals can use a range of $10M/mile to $15M/mile for all financial costs of putting a PRT system into operation.
Learn more about PRT at www.electric-bikes.com/prt-info.html
