Alternative PRT Systems for Kungens Kurva
This document is part of the Local Assessment Plan for the Swedish part of the EDICT (Evaluation and Demonstration of Innovative City Transport) project financed by the EC and Vinnova
by Dr. Ingmar Andréasson
Revised January 25, 2003
7. Alternative PRT systems for Kungens Kurva
The EDICT project is based on the ULTra PRT system as demonstrated on the full-scale test track in Cardiff. The fact that ULTra has a running test installation puts this system in the lead with respect to proven technology and availability for public operation.
This is a chapter from a report within the Swedish part of the EDICT project. EDICT is the acronym for Evaluation and Demonstration of Innovative City Transport. This project is sponsored by the European Commission to assess the impacts of PRT (Personal Rapid Transit) in four different sites in Europé (Cardiff, Eindhoven, Ciampino-Rome and Huddinge-Stockholm). As part of the assessment for the Swedish site we have made an evaluation of some alternative PRT systems.
The Swedish site is Kungens Kurva (Kings bend) in the city of Huddinge adjacent to Stockholm. Kungens Kurva is a shopping and working area with 2300 employees and 34 000 daily visitors, the area being poorly served by public transport. We are evaluating a PRT system with 18 stations, 12 kms of guideway and some 90 vehicles to serve this area at the present level of demand.
There are several other developments of PRT systems around the world. One PRT-like system has been running in Morgantown USA since 27 years. The Morgantown vehicles are larger than PRT (8 seated and 12 standing passengers) but as for PRT all stations are off-line and vehicles only visit stations on demand. With 1970 computers and software they could operate at 7,5 seconds headway although they did operate at 15 second headways. Both computers and software have been updated very recently.
A recent survey made by the Advanced Transit Association (not yet published) lists 23 and describes 15 PRT systems in various stages of development.
We have chosen to compare ULTra with two other systems, viz. Taxi 2000 and Austrans. This selection was based mainly on their degree of development. Taxi 2000 is probably the most technically advanced and solidly researched system while Austrans has better organisational resources and a vehicle running on a test track. Another reason to include Austrans is that it can be viewed as a small-scale LRT and an LRT route is planned to pass by or through the Kungens Kurva area. Austrans vehicles are larger than PRT and more suitable for GRT (Group Rapid Transit) although they can operate as PRT during off-peak hours.
Among the PRT systems not analysed for this application at least two stand out as being potentially relevant. MicroRail is as far developed as Taxi 2000 in terms of hardware prototypes of vehicle and guideway but probably not as advanced in terms of software and control. SkyCab is a Swedish industry consortium with alleged financing for development of a prototype and potential applications in the Stockholm area. SkyCab has chosen not to provide information on their development status.
After describing each system we compare properties and discuss the differences. We do not aim at presenting a single winner but to review properties in a way that can be used as a decision support for a specific application. As of November, 2002, there was no market-ready PRT system to choose from. Being available is the first and necessary criterion. The second criterion may be that the technology has also been proven in public service. The assessment of a PRT system in operation will be an important contribution of the EDICT project.
7.1. Taxi 2000
A multi-disciplinary research team at the University of Minnesota developed the Taxi 2000 PRT system starting already in 1982. Professor J. Edward Anderson is today the leading expert on PRT systems. The university has licensed the rights to Taxi 2000 Corporation lead by Anderson. Taxi 2000 in turn licensed the rights to Raytheon Corporation who changed the specifications and named their version PRT 2000. Raytheon demonstrated on their test track the technology, control and safety with 3 vehicles running at 48 kph with 2,5 seconds headway.
Vehicles and guideway in Raytheon's version were large, heavy and expensive. When Chicago RTA did not accept the price Raytheon chose to discontinue the development. All rights have been returned to Taxi 2000 who have a development team designing and building a prototype.
Figure 7.1. Scale model of Taxi 2000 Figure 7.2. Rendering of Taxi 2000 on guideway
Taxi 2000 vehicles seat 3 adult passengers (side by side) or alternatively 4-5 children or one wheelchair (facing sideways) or baby-carriage with attendant. Vehicles run on smooth rubber tires in a U-shaped trough. Propulsion is electric with two linear induction motors (LIM) fed from a power rail with battery backup on board. Switching is mechanic with a switch arm in the vehicle engaging in guide rails. Speed is 32-65 kph. Initial testing is with 3 seconds headway.
The control system is decentralized asynchronous with critical components and software doubled for redundancy.
Taxi 2000 claim to have invested 32 million in development, including software for simulation, operation and control. Recently 0,5 million were raised to build a full-scale prototype and a piece of guideway. The official inauguration of the prototype is planned for the end of January 2003.
Figure7.3 Rendering of Taxi 2000 prototype now under construction
Figure 7.4. ULTra prototype vehicle
ULTra (Urban Light Traffic) is a PRT system developed by a research group under the leadership of Professor Martin M. Lowson at the University of Bristol, UK. Battery powered vehicles run on rubber tires on concrete/steel guideways elevated or at grade. Instead of being guided and switched mechanically, vehicles navigate using on-board maps, wheel rotations, wheel angles and magnetic markers in the guideway. Steering is by the front wheels and guideway sides act as structural members and safety barriers. Batteries are charged (or exchanged) during standstill at stations or in a depot.
ULTra has been dimensioned for 2+2 passengers facing each other and running at 40 kph.
Thanks to navigation and battery power the guideway can be made simple without guide-rails and power rails.
The control system is centralized and synchronous. Before a vehicle is allowed on the main track it has been allocated time-slots for passage through all conflict points.
Ultra has been developed since 1995 with a total investment of 10 million so far.
Figure 7.5. Two ULTra vehicles running on the test track.
Two prototype vehicles and a 1 km test track are in operation. The city and region of Cardiff in Wales have decision and funds budgeted for the first commercial PRT installation to operate in 2005. The initial stage has 3,3 kms of guideway, 5 stations and some 40 vehicles.
Figure 7.6. Photomontage of ULTra guideway in Cardiff
Figure 7.7. Austrans prototype vehicle
Austrans is based on larger vehicles than pure PRT systems and is classified as Group Rapid Transit (GRT). Austrans vehicles accommodate 9 passengers, all seated. Austrans can be operated as individual PRT except in peak hours when ride-sharing is applied and vehicles make intermediate stops. All stations are off-line so that stopping vehicles do not block the main line.
Austrans wheels can grip around special Z-shaped rails permitting steep grades and safe braking. Wheel arrangements allow sharp turns down to 8 meters which is unique for a rail vehicle (but not for PRT). Rails are flexible (!) so that they can be switched very fast (1 second).
Another unique feature of Austrans is lateral movement of rail segments at stations. Loading and unloading of vehicles can then be made independently, thus increasing station capacity.
Austrans have invested 5 million so far and have secured financing of another 15 million for developing vehicles and test track. Two vehicles are running on a test track including a chicane curve and a switch. The Bishops Group developing Austrans is large and well renowned as designers and suppliers of wheel suspension/steering for cars.
The performance of Austrans positions that system above PRT both with respect to speed (up to 120 kph intercity) and capacity. Large vehicles compensate for the long headway imposed by the rail switch.
Figure 7.8. Austrans test track outside Sydney
7.4. System parameters
Important physical parameter are summarized in table 7.1 and commented on below.
Guideway W x H m
|Curve radius m||Grade Climbing %||Support spacing, m||Vehicle dim., L x W x H, m||Vehicle weight, kg|
|Taxi 2000||steel frames covered U-shaped||mechanical against guide rail||2 lims, power rail||1.0 x 1.0||12||10-15||29||2.6 x 1.4 x 1.5||455|
|ULTra||concrete/steel open||steered wheels navigation||1 rotary motor, battery||2.0 x 0.45||6||15||18||3.7 x 1.5 x 1.6||400|
|Austrans||steel rails for grip wheels||flexible rails||2 rotary motors, power rail||1.9 x 1.1||8||20||20||5.4 x 1.9 x 2.3||2125|
All systems have vehicles rolling on a runway. ULTra has an open concrete guideway with steel sidings. Austrans rails need a concrete bed for support. Taxi 2000 has a steel space-frame guideway covered with panels leaving a narrow upward slot. Both ULTra and Taxi 2000 have downward openings for drainage of rain and snow. ULTra openings are wide enough to let through the sunlight. Taxi 2000 and ULTra run on rubber tyres while Austrans use steel wheels on steel rails.
The normal PRT solution is for the vehicles to switch and the guideway to be fixed. Austrans has a rail switch requiring a longer time headway. ULTra has no switching mechanism but steers through switch areas with the front wheels.
ULTra and Austrans like most other PRT systems are propelled by electric rotary motors. Taxi 2000 uses linear induction motors (LIM). Taxi 2000 and Austrans have double motors for redundancy.
Ride comfort is normally the limiting factor for curve radii. Taxi 2000 state that less than 12 meters radius would necessitate impractically low speeds. The 6 m radius of ULTra is probably only used in stations. The 8 m radius curves of Austrans are negotiated at 17 kph.
Taxi 2000 can climb 10-15 % grade depending on motor size. Austrans can handle as much as 20 % grade allowing steep ramps to stations at ground level. In practice the limit on grade is probably comfort requirements of passengers.
Spacing of supports
A practical limitation on support spacing is the length of guideway sections during transport to the construction site.
|System||Power kW||Noise dBA||Pass.cap.||Headway secs||Link cap., pph||Speed, kph||Guideway cost, M euros/km||Vehicle cost, K euros||System cost, M euros/km|
|Taxi 2000||4||n.a.||3||0.5 - 3||3,800||32 - 65||1.6||26||3.5|
|ULTra||2||42||4||1 - 3||3,300||40||2||50 (prototype)||4.5 - 9.1|
|Austrans||39||77||9||2.5 - 4.2||7,600||17 - 70||5.4||131||9|
Table 7.2. Performance parameters
Power consumption for ULTra does not include power for heating or A/C (not foreseen).
We have no figure for the noise from Taxi 2000 vehicles. The fact that they run on slick rubber wheels enclosed in a covered guideway leads us to believe that Taxi 2000 noise is lower than that of ULTra who use textured tyres on an open guideway.
Vehicle capacities over 4 passengers make sense only with ride-sharing.
The smallest safe headway between vehicles is limited by speed, reaction time, braking power and comfort requirements. A common requirement is the so called brickwall safety imposing that if one vehicle would somehow come to a sudden stop then the following vehicle must be able to stop in time to avoid a collision. In most cases the comfort limitation on emergency braking is the one parameter that limits the headway. Hence the time headway is essentially determined by speed and is independent of system. Safety belt or airbag would allow shorter time headways.
Taxi 2000 and ULTRa start testing with 3 second headways and intend to reduce headways when the reliability has been verified. The 4,2 seconds headway for Austrans is required to throw the rail switch, to verify the switch position and to stop before the switch if necessary. On links the headway can be reduced to 2,5 seconds at 70 kph.
Link capacity is calculated from vehicle size and time headway with 30 % empty vehicles and all loaded vehicles filled to capacity. ULTra capacity was further reduced to 65 % since not all time-slots can be occupied. We assumed 2 seconds headway for both Taxi 2000 and ULTra and 3 seconds for Austrans. Practical capacity is lower (about half) because vehicles are not fully loaded.
In a large PRT network the traffic load can be distributed over several links so that network capacity is higher than link capacity.
In many systems guideways can be designed for varying speeds. Higher speeds require wider curves but they offer shorter trip times and smaller fleet size. High speeds require longer time headways and hence reduce link capacity. Austrans can run at 120 kph outside cities.
Note that effective travel speed from origin to destination equals line speed when vehicles do not stop en route.
Guideway costs include suppots, foundations, control and power. ULTra vehicle cost is the actual cost of the prototype. System cost is presented as a typical total cost per system kilometer of single guideway including stations, vehicles and control system. The low figure for ULTra is for low demand (few vehicles) and 25 % elevated guideway whereas the high figure is for high demand and 75 % elevated guideway. The other systems assume 100 % elevated guideway.
In comparison a new LRT line in Sweden costed 14-19 million per kilometer double track.
7.5. Distinguishing features
The larger vehicles of Austrans presuppose ride-sharing and some stopping en route (during peak hours) in order to be reasonably filled.
Austrans can go at high speed (120 kph) and is therefore suitable for longer trips. In urban areas 70 kph is planned and in curves as low as 17 kph.
Austrans is the only system to use a rail switch. A rail switch takes time to throw and approaching vehicles must be able to stop before the switch if for some reason it should be in the wrong position. Guideway switches mean longer headways reducing capacity. They will require some maintenance at least to prevent them from freezing.
Stations at grade
Stations at grade is an advantage (low cost and good accessability) but it takes power to brake downhill and to accelerate uphill. Austrans can descend to ground level via short and steep ramps. We think that ground level stations are practicable only where the main guideway is at grade or where elevations in the landscape can be used (or constructed).
The ULTra vehicles can be further developed to run on ordinary streets in mixed traffic. Hence ULTra has the potential of developing into a dual-mode system offering non-stop travel from door to door with privately owned (or rented) vehicles.
ULTra is planned for battery operation with battery charging or exchange during standstill. Charge time is estimated to 20 % of run time. Battery range is claimed to be around 5 (short) trips.
In the (rare) event of a breakdown a PRT vehicle is normally pushed by another vehicle to the nearest station. If it cannot be pushed it has to be taken off the guideway with a sky-lift. Only incase of a fire on board should passengers be allowed to leave a vehicle on the guideway (and stay there until picked up by a sky-lift).
ULTra has a hatch in the front window for emergency egress. The guideway has a walkable grid between the running surfaces. There is no dangerous voltage on the ULTra guideway.
This illustration does not reflect that Taxi 2000 has a longer guideway span and hence fewer supports.
7.6. Comparative assessment
I have made my subjective assessment of relative strengths and weaknesses of each system. The results are summarized in Table 7.3 and commented on thereafter. Better than average or normal is represented by + and below average by --. No mark is normal.
|Stations at grade||--|
Table 7.3. Pros and cons
Austrans guideway and vehicles are large than the other two.
Stations at grade
The guideway of Taxi 2000 is impractical to put on ground level (1 meter high).
Austrans with its steel wheels and steel rails makes much more noise than the other systems.
Austrans uses much more energy than the other systems both on a per vehicle basis and per passenger.
Most PRT and GRT systems use mechanical guidance by rails or curbs. ULTra vehicles navigate and steer freely with the front wheels. The choice has several advantages and a few drawbacks:
+ Can run on flat surface without curbs
+ Less demand on guideway alignment
+ Can develop into Dual-mode
+ Can bypass within stations
Steering relies on friction
Bypassing (if allowed) requires larger stations
With free navigation it is possible to store empty (and disabilitated) vehicles in stations without blocking other vehicles. Bypassing in stations will reduce waiting when vehicles ahead are not ready to depart. On the negative side, the flexibility of bypassing requires additional maneouvring space as shown in figure 7.10.
Figure 7.10. Linear stations as used with mechanical
guidance and independent loading berths as possible
with free navigation. Vehicle dimensions correspond
to Taxi 2000 and ULTra respectively.
It is often claimed that station capacity is the bottleneck of a PRT system. It may be so but then it is due to incorrect dimensioning. It is true that longer platforms bring decreasing additional benefits and it does not help much to go beyond 7 or 8 loading positions without bypassing possibilities. Another way to handle higher demand is by building parallel platforms or split up a station into two.
ULTra and Taxi 2000 vehicles are probably equivalent in terms of loading/unloading time. ULTra can take one more passenger but all four seats will seldom be used. Taxi 2000 has one bench where ULTRa has two but we think that passengers pass the door one-by-one. ULTra vehicles can move as soon as they are loaded while a Taxi 2000 vehicle has to wait if the vehicle in front is not ready to move. On the other hand ULTra vehicles have to wait longer before the whole path is free so that they are allowed on to the main track. Hence ULTra station capacity is higher per platform position but maybe not higher per station area unit.
Austrans vehicles are larger so that loading time per passenger is lower. For a given station size we estimate that Austrans station capacity is higher.
ULTra plans to use synchronous control at least initially. Vehicles on the main tracks follow virtual slots moving at the same speed. Slots for passage through merge points throughout the whole trip are booked in advance before a vehicle is allowed to leave a station. In order to find synchronized free slots on a long trip some 35 % of all slots need to be kept free.(reducing link capacity) All bookings are kept in a central computer and as long as all vehicles maintain their speed there will be no conflicts. For added safety each vehicle checks the distance to the vehicle ahead. Vehicles approaching the same merge point need to be checked by a local merge controller.
Most other PRT and GRT systems are based on asynchronous control (or a variation called point-synchronous control). Vehicles may leave a station whenever there is space on the main track. When a vehicle enters a link leading to a diverge it calls the node computer for (dynamic) route choice. When approaching a merge it calls that node processor to be assigned a passage time. Hence processors are needed to handle each node (one processor can handle several nodes) and two-way exchange of messages is necessary on each link.
The choice of control system has several implications on cost, capacity, flexibility and reliability. The advantages of each system are summarized as follows:
Synchronous control Asynchronous control
+ fewer processors in network + decentralized control less vulnerable
+ little communication + vehicle failure only affects one link
+ low cost + route change is possible
+ no congestion in network + less waiting to depart
+ 100 % of link capacity can be used
+ less empty running
+ empty vehicles can form platoons
Taxi 2000 is independent of road friction thanks to magnetic propulsion and braking.
The open profile of ULTra requires snowplows on vehicles, special sweeping vehicles and chemical melting. As a safety measure it is possible to monitor guideway friction continuously and reduce speed or increase headways globally as needed. During extreme weather conditions, operation may have to be temporarily stopped.
ULTra vehicles can be evacuated on the guideway.
Taxi 2000 is claimed to cost less than half compared to the other systems!
Only ULTra has the potential to develop into a Dual-mode system. Devices for manual control would have to be added and one seat would probably have to be sacrificed.
ULTra and Austrans already have vehicles running on a test track. Political decisions and funding are in place to have a PRT system operating in Cardiff with passengers during 2005. Taxi 2000 is two months from having a test vehicle on a short guideway segment.
ULTra provides a simple solution, close to market availability. However, it is not weather resistant enough for the Scandinavian climate without amendments. The synchronous control strategy can work in small networks and can be replaced later to handle larger networks.
Taxi 2000 is the most engineered PRT system with the lowest cost and the most weather-proof design. They need additional funding before being available on the market.
Austrans, being both PRT and GRT is heavier and more costly, but, on the other hand offers combined PRT/GRT service at a higher speed with a higher capacity. Austrans with its industrial backing seems to be close to market readiness.
For implementation of a PRT system now in Kungens Kurva we recommend ULTra with facilities for snow clearing and a monitoring system for guideway friction. Austrans is recommended for a local system in Kungens Kurva if it is accepted to replace the planned LRT system passing the area. Our recommendation would change in favour of Taxi 2000 if it became market-ready before a decision is taken.
Last modified: January 27, 2003