PRT Costs Compared to Bus, LRT and

Heavy Rail –Some Recent Findings

Paper to be presented at:

AATS European Conference in Bologna 7-8 Nov, 2005
”Advanced automated transit systems
designed to out-perform the car”
by Göran Tegnér, , M Pol.Sc.
TRANSEK Consultants,
Sundbybergsvägen 1A, SE 171 73 SOLNA, Sweden
goran@transek.se Mobile phone: +46-707-21 30 74
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

2 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

CONTENTS
1 ABSTRACT...............................................................................5

2 AIM OF THE STUDY ................................................................5

3 AVAILABLE DATA ....................................................................6

3.1 PRT Cost Data..........................................................................6
3.2 Traditional public transport costs ..............................................6

4 A TRANSIT COST MODEL ......................................................7

4.1 Model assumptions ...................................................................7
4.2 Passenger capacitities ..............................................................8
4.3 Model results ............................................................................9
4.4 User costs ...............................................................................11
4.5 A Comparative Example .........................................................13
4.6 Sensitivity analysis..................................................................15
4.7 PRT compared to traditional transit ........................................17

5 CONCLUSIONS AND REFLEXIONS .....................................19

5.1 Conclusions ............................................................................19
5.2 A personal reflexion ................................................................20

REFERENCES................................................................................22

6 ACKNOWLEDGEMENTS .......................................................23

APPENDIX 1: PRT COSTS ............................................................24

Capital costs....................................................................................24
PRT Capital cost comparison..........................................................26
PRT Operating Costs ......................................................................27

APPENDIX 2. COST COMPARISON .............................................30

Capital costs for guideway ..............................................................30
Capital costs for vehicles ................................................................31
Economic life time and discount rate ..............................................32
Operating costs ...............................................................................33

APPENDIX 3. OTHER COMPARATIVE COST DATA....................35

3 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

1 ABSTRACT

In the EDICT project, a comprehensive social benefit-cost analysis was carried out as regards the
merits of a Personal Rapid Transit System for the Kungens Kurva (“King’s Curve”) Area in Huddinge,
south-west of Stockholm, Sweden. As a part of this project, the costs of the ULTra PRT system was
compared to a few other prototype PRT systems, such as the Taxi 2000 system (Skyweb Express) and
the Austrans Group Rapid Transit.

In a recent project, financed by the Stockholm Chamber of Commerce, we compared Capital and
Operating Costs on a per passenger-kilometre basis between Bus, LRT, Metro and Commuter Rail
services in Stockholm.

In this paper, the cost comparisons have been broadened to include PRT as well. This means that the
cost structure of PRT can be compared to the costs of the Bus, the LRT and the two heavy rail modes.
Consideration has been taken to capacity limits, the economic life time of various components, such as
the guideway, stations and vehicles.

A cost model has been developed in order to make it possible to compare various transport modes in a
consistent way and we have elaborated the cost data from recent studies (see appendices) and applied
in the model.

The major findings are the following ones:

• PRT is the cheapest mode of urban public transport in a wide range of passenger demand
• The bus mode is the second cheapest mode, but it has a limited capacity
• LRT is a cheaper mode than the heavy rail modes up to some 25,000 passengers per day (in
both directions)
• Over the full practical range of demand levels, the total capital plus operating costs per
passenger-kilometre for the PRT mode is less than 1/3 of a LRT system and also cheaper than
Bus in the lower range, relevant for the bus mode.

PRT is therefore a most cost-efficient mode of urban transport. This paper will summarize these
results.

2 AIM OF THE STUDY

PRT is a brand new mode of (urban) public transport system, not yet in full operation anywhere. After
some 40 years since it was originally launched in the mid-1960, there are still just a couple of test
tracks for prototype systems. The pretended merit of PRT is always questioned by the transport
authorities almost everywhere.

“If it is so fantastic, why hasn’t PRT been implemented yet somewhere”, is a typical question.

5 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

“It seems promising, but the costs must be high or at least very uncertain”, is another typical question.

The aim of this study is to investigate into the cost of PRT and to compare PRT costs to the costs of
traditional urban public transport, such as the Bus, the LRT, the Metro and the Commuter Rail modes,
respectively. We will examine both the capital cost of the infrastructure, the capital costs for vehicles
and the operating and maintenance costs. However, the costs for traditional transit modes are restricted
to empirical data for the Metropolitan Stockholm Area. All costs are in 2002-2003 price level.

3 AVAILABLE DATA

3.1 PRT Cost Data

PRT cost data is in a sense not quite comparable to the costs of the traditional public transport modes,
such as the bus, LRT and heavy rail modes, as PRT is not yet in full operation anywhere.
Nevertheless, there are many impressive efforts from PRT developers and others that have undertaken
serious attempts to calculate the PRT costs very careful. The data sources used in this brief
comparison are the following ones:
• ULTra PRT costs obtained from prof. Marin Lowson at Advanced Transit Systems Ltd.
(investment and operating cost model)
• ULTra costs obtained form the EDICT project
• ULTra costs from the comparison between calculated PRT costs for Cardiff compared to
tendered PRT costs from two independent building companies (refers to the guideway costs)
• Skyweb Express costs obtained from PSrof. Edward J. Andersson at Taxi2000
• Austrans costs obtained from Austrans
• General PRT Costs obtained from ATRA

As Austrans is a Group Rapid Transit rather than a pure PRT system, Austrans cost would yield an
upper cost boundary for PRT costs.

3.2 Traditional public transport costs

The cost for the traditional urban public transport modes are in general based upon metropolitan
Stockholm data. The public transport authority in the county if Stockholm, the Greater Stockholm
Transit Company (AB-Stor-Stockholms Lokaltrafik, SL) have excellent records. Investment cost data
are derived from recent LRT and Commuter rail projects in the Stockholm area. Investment costs for
Guided bus are derived from the Zuidtangent project in Amsterdam (42 km). LRT Cost is also
compared to several other European cities.

6 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4 A TRANSIT COST MODEL

In this chapter we present a Transit Cost model, where we use the information presented in earlier
studies, presented in appndices. First the underlying assumptions are briefly presented, and second the
resulting model is presented. In a third section a comparative example is given when PRT and LRT is
compared, with the aim to show how much transit one gets at the same amount of investment capital.
In a fourth section a sensitivity analysis is presented.

4.1 Model assumptions

The following model assumptions form the basis for the comparative cost model:

• All capital costs are converted into an annual capital cost, using an annuity formula
• The annuity is calculated a 4 % real discount rate and according the expected life time of the
various types of investment (track, vehicle, station, depot etc.)
• From the capital cost for the infrastructure, the capital cost for vehicles are deducted. The
remaining capital costs (for tracks and stations) are regarded as fixed and independent of the
number of daily passengers, while the vehicle costs area assumed to be dependent on the level
of ridership.
• A minimum and a maximum range for the passenger capacity is calculated for each type of
transit system. For the heavy rail modes it is based on the number of cars per train
• A minimum service level of a 15 minute headway is assumed
• A maximum service level of 5 minute headway is assumed for the traditional transit modes
such as bus, LRT and the heavy rail modes. The theoretical minimum time between departures
might be somewhat lower, say 2 minutes, but then passenger loading and unloading times at
stations will be restricting the actual practical capacity.
• For the separate BusWay a 2,5 minute headway has been allowed
• For PRT 2 second headway is assumed to be the maximum service level (and 5 minutes as a
minimum level).
• The peak demand is assumed to amount 10 % of the daily demand
• At the maximum service level the practical passenger capacity is assumed to amount 85 % of
the total theoretical, due to an uneven distribution of arrivals
• All costs (annual capital costs for the guideway and stations, for the vehicles and the operating
and maintenance costs) are calculated at a 10 kilometre standard trip.

7 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4.2 Passenger capacitities

The passenger capacity at minimum and maximum service levels is presented in Table 1 below:
Table 1. Vehicle and daily passenger capacity for various transit modes according to assumptions
Daily
Daily pass.cap. pass.cap. at
Cap./veh.(seated Vehicles per Vehicles per Passenger/ Passenger/ at min service maxn service
Transit mode & stand) train (min) train (max) train (min) train (max) (2 directions) (2 directions)
City Bus 119 1 1 119 119 9 520 24 276
Bus Way 119 1 1 119 119 9 520 24 276
LRT 106 2 6 212 636 16 960 129 744
Metro 138 6 9 828 1 242 66 240 253 368
Commuter rail 150 6 8 900 1 200 72 000 244 800
AGT 106 2 6 212 636 16 960 129 744
PRT 4 1 1 4 4 57 600 122 400

The daily capacity for the analyzed transit modes is also presented in Figure 1 below:

Figure 1. Daily passenger capacity for various transit modes at minimum and maximum service levels

Daily passenger capacity at minimum and maximum

service level for various transit modes

275 000
250 000 Daily pass.cap. at min service (2
directions)
225 000 Daily pass.cap. at max service (2
directions)
200 000
175 000
150 000
125 000
100 000
75 000
50 000
25 000
-
City Bus Bus Way LRT Metro Commuter AGT PRT
rail

The vehicle passenger capacity figures for the traditional transit modes are derived from the Greater
Stockholm Transit (SL) Company. PRT yields a fairly high daily capacity in spite of its small vehicles.
This is due to its extremely high frequency of service.

8 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4.3 Model results

The cost model is calculated in the range from 5,000 and up to 250,000 daily trips.

To summarize the final findings of this research task, one might conclude that PRT is the cheapest
urban public transport mode among the studied ones, such as Bus, LRT, Metro and Commuter rail, see
Figure 2 below:

Figure 2. Annual Investment and O & M costs per passenger-trip (10 km) for various transit modes

Total cost per passenger-journey

(10 km) for various transit modes
€ per Passenger journey (2 directions)
10
City Street Bus
9
Bus Way
8 LRT
LRT
Metro
7
Commuter rail
6 AGT
Commuter rail PRT
5

4
Bus Way AGT
3
Metro
Street Bus
2

1 PRT

-
5 000 10 000 15 000 20 000 25 000 30 000 35 000 40 000 45 000 50 000

In Figure 2 above, both investment and operating and maintenance costs are compared. A standardized
urban journey of 10 kilometres has been used for the comparison.

The average cost per passenger is € 2,50 for the street bus in its capacity intervall up to 25,000 daily
trips. All other modes show declining average costs as ridership grows

The BusWay mode (run on bus-only lanes) is defined to have a maximum capacity of 50,000
passengers per day and both directions and within this limit, the BusWay mode is the cheapest of all
the traditional modes above a ridership of 10,000 journeys. The overall cost per passenger amounts
about € 2,5 per 10-kilometre trip for bus (with a variation from € 5,5 down to € 2 per 10 km trip for
the BusWay). The main reason for the low bus cost is the lack of infrastructure cost for the guideway.
The only relevant costs are the capital cost for vehicles and the operating and maintenance costs (O &
M costs). Here we assumed that for the urban standard city bus, the road network is already built,

9 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

mainly for car traffic. The Bus Way reflects the costs for a bus-only street corridor, such as the
Zuidtangent Bus lane in Amsterdam.

The LRT mode is the second cheapest traditional urban mode up to the same passenger load, 25,000
passengers per day (both directions). The average costs for LRT varies from € 7,1 per trip down to €
4,12 per trip (the latter at 25,000 passengers). At 25,000 passengers per day, LRT is about 64 % more
expensive per passenger trip than the city street bus.

The two heavy rail modes, Metro and Commuter rail show a rather similar cost pattern. These two
systems are built to handle high passenger volumes, even much higher than the 50,000 passengers per
day shown in Figure 2 above. At higher loads, say from 25,000 passengers per day, the commuter rail
is 4 % cheaper then the metro system, € 4 compared to € 4,12 per trip. At even higher loads than
25,000 passengers per day, commuter rail becomes even cheaper than the metro system.

At a load of about 30,000 passenger journeys per day, the metro and commuter rail modes become
cheaper than the LRT-mode. At 50,000 passengers per day, the average cost is € 3 for the metro and €
2,60 for the commuter rail, compared to € 3,75 for the LRT per 10 km trip. Metro and rail is thus 20 %
- 30 % cheaper than LRT at this demand level. The average cost per trip falls sharply for the two
heavy modes, as ridership increases, due to economics of scale. The explanation for this is that the
metro and rail investment costs are very high, while the operating costs are very low per passengers
compared to the LRT mode.

In Figure 3 below a variant of the cost model is shown, with respect to the minimum level of service,
i.e. at least 15 minute headway for all traditional modes, and 5 minute headway for the PRT mode.

10 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 3. Annual Investment and O & M costs per passenger-trip (10 km) with respect to minimum
service (15 min headway) for various transit modes

Total (producer) cost per passenger-journey

(10 km) for various transit modes
€ per Passenger journey
10

City Street Bus

9
Bus Way
8 LRT
LRT Metro Metro
Commuter rail
7
Commuter rail
6 AGT
PRT
5

4
Bus Way AGT
3
Street Bus
2

1 PRT

-
5 000 10 000 15 000 20 000 25 000 30 000 35 000 40 000 45 000 50 000

In this case, the cost per passenger-journey at the lower ridership levels becomes higher, as the
capacity necessary to provide the minimum service will exceed the travel demand. The average cost
functions thus become steeper, especially for the heavy modes metro and rail but also for LRT. In this
case LRT becomes more competitive towards the two heavy rail modes, with lower LRT cost per
passenger-journey up to 45,000 trips per day.

PRT is still the cheapest mode, with about half the cost per passenger-journey compared to the
BusWay at 25,000 passengers per day, and less than 1/3 of the LRT cost.

4.4 User costs

Another aspect of the total cost srructure is the cost per trip, that the traveller beears himself/herself.
These so called user costs, usuaölöy consist of the follwing travel time components and of a fare:
• Walking time
• Waiting time
• In-vehicle time
• (and sometimes also: Transfer time)
• Fare

Form the Metropolitan Stockholm Area, Table 2 is derived with realistic averages for travel time com-
ponents, headway and fare level:

11 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Table 2. Travel time components and generalized time and cost for traditional and PRT modes in
Stockholm

Time components in minutes Generalized Generalized Generalized

Mode of transport Walk Headway Wait In-vehicle Total time Time, min Cost,€ /10 km trip
User costs per 10-km trip incl. a 2 € fare
Bus 15 30 15 40 100,0 100 12,6
LRT 5 10 5 24 44,0 44 6,7
Metro 10 4 2 14 29,6 38 6,0
Commuter rail 15 15 7,5 12 49,5 57 8,1
PRT 5 <1 0,5 17 22,2 29 5,0

Travel time weight 2 2 1

Travel time value, € per hour 6,4

Source: Stockholm Data and own calculations

The generalized tiome is calculated with the weight of two (2) for walk and wait time, and the
generalized cost is calculated with an average travel time value of € 6,40 per hour. A € 2 fare per trip
has been assumed equal for all transit modes.

As can be seen from Table 2, the PRT mode shows the lowest user cost per trip of all modes.
Compared to the Bus mode, PRT yields only 60 % of the total generalized cost. The Metro mode has a
19 % higher Generalized Cost compared to the PRT mode. The generalized cost thus includes the
weighted travel time components and the fare. When combining both the producer cost and the user
cost into a total cost per 10 km trip, a quite different picture shows up:

Total (producer & User) cost per passenger-journey

(10 km) for various transit modes
€ per Passenger journey
40
Bus Way
City Street Bus
35 Bus Way
LRT
Commuter rail Metro
30
Commuter rail
25 Metro AGT
PRT
20
Street Bus LRT

15 AGT

10

PRT
5

-
5 000 10 000 15 000 20 000 25 000 30 000 35 000 40 000 45 000 50 000

Compared to the bus mode, the total producer and user cost for PRT is 55 % lower; compared to LRT
PRT is 40 % cheaper, and compared to the heavy rail modes, PRT costs are 50 % and 56 % cheaper
respectively. For the bus mode, user cost is 5 times higher than the producer cost. Fore heavy modes
they are fairly equal, while for LRT, PRT and AGT user costs are between 1.6 and 3.2 times higher
than the producer costs. PRT is the cheapest mode also when the user costs are included.

12 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4.5 A Comparative Example

Assume a budget for public transport investment of 125 mUS$, or 104 m€. Let us compare a typical
LRT corridor line with a PRT loop system. What could be obtained for the same amount of investment
money?

A) A LRT System
With a LRT system the optimal station spacing is about 700 meters.
Figure 4. A LRT line for 125 mUS$

A LRT System – investment cost 125 mUS$:

5 km track & 8 stations:

The main LRT results would be:

• An area within 600 meter of walking distance: 9 sq.km
• Number of served inhabitants: 23 500
• Practical headway: 5 minutes (2-3 minutes theoretical)
• Average door-to-door speed: 15 km/hour (incl. walk and wait time)
• Travel time for a 5 kilometre ride: 20 minutes without road congestion
• Hourly passenger capacity at 5 minutes headway: 2 544 pass

B) A PRT system

A PRT would be designed in loops rather in a corridor fashion. For comparative reasons, we have kept
the same station spacing as for the LRT system, to yield the same walking distance.
However, with PRT the optimal stations spacing is rather 250 meters.

13 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 5. A PRT network for 125 mUS$

A PRT-system - Investment cost: 125 mUS$:

20 km track & 33 stations

The main PRT results would be:

• An area within 600 meter of walking distance: 37 sq.km
• Number of served inhabitants: 97 000
• Practical headway: 5 seconds (theoretical: 2 seconds)
• Average door-to-door speed:23 km/hour (incl. walk and wait time)
• Travel time for a 5 kilometre ride: 13 minutes irrespective of road congestion
• Hourly passenger capacity at 5 seconds headway: 2 880 pass

14 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 6. Performance comparison between LRT and PRT.

Performance comparison – LRT ÙPRT

120
97
100

80

60

40
20 23,5
20 13

0
Trip time 5 km, min Residents served, 1000's

LRT PRT

The conclusion is therefore, that the served area, and, consequently, the number of served residents,
would be four times higher with PRT compared to the LRT system. Also, the door-to-door travel time
might be roughly only half (or slightly more) with PRT compared to LRT. This is not only due to the
much shorter waiting times, but also attributable to the total absence of necessary deceleration, stand
still and acceleration times at all the intermediate stops ‘en route’ with the PRT, with its off-line
stations1.

4.6 Sensitivity analysis

How competitive will the PRT mode be compared to the other, more traditional transit modes, if we
change the economic life time for the PRT guideway, stations and vehicles? A PRT is an automatic
light mode, and cheaper in construction compared to the heavier modes, one might argue that the
duration of the fixed and roiling stock equipment might be substantially shorter. As Taxi 2000 does,
let us assume the following life time for the various types of equipments:

1
See M. Lowson, ATS Ltd: ”Service Effectiveness of PRT vs. Collective Corridor Transport”. EDICT,
European Commission,. Technical Note ATS TN 2002-10. Document date: 2002-11-29

15 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Table 3. Assumed Economic Life Time of PRT guideway, stations and vehicles

Type of equipment Economic life time in years Economic life time in years
“Standard assumption” “Sensitivity analysis”
Guideway 60 30
Stations 25 10
Vehicles 25 8

With the shorter economic life time, the capital costs for PRT will rise.

Figure 7. Sensitivity analysis: Cost model comparison for various transit modes, with shorter Life time
for PRT equipment

Sensitivity analyis with shorter life time for PRT

Total cost per passenger-journey
(10 km) for various transit modes
€ per Passenger journey
10

9 City Street Bus

Bus Way
8 LRT
LRT
Metro
7
Commuter rail
6 AGT
Commuter rail PRT
5

4
Bus Way AGT
3
Metro
Street Bus
2

1 PRT

-
5 000 10 000 15 000 20 000 25 000 30 000 35 000 40 000 45 000 50 000

The total (investment and O & M) costs for the PRT mode now augments by 33 % from € 1,19 up to €
1,59 per 10 km passenger journey.

Only at the lowest ridership level, 5,000 trips per day, will the ordinary Street Bus and PRT show a
similar cost level. And the level of service will still be very different in terms of waiting and in-vehicle
travel time.

Still, at the 25,000 demand level, the PRT cost per passenger-journey will become 25 % cheaper than
the BusWay, 37 % cheaper than the Street Bus and 62 % cheaper than the LRT mode. The LRT mode
is 2,6 times more expensive than PRT.

16 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

4.7 PRT compared to traditional transit

The PRT mode is the cheapest mode of all the five compared public transport modes. This is because
the investment costs for both the track and the vehicles are low due to the light weight of these PRT
components, and also that operating costs are low, due to the driverless automation of the PRT system.
PRT costs falls from € 2 at very low passengers loads (5,000 trips per day) down to € 1,1 per 10 km
trip at the higher loads.

PRT is the cheapest of all traditional public transport modes. Even if the PRT cost would double
compared to today’s estimates, if would still become cheaper than both the City Street Bus, the LRT
and the heavy modes, Metro and commuter rail. It is only the Bus Way with its € 2 per trip that would
be able to compete with the doubled PRT costs.

Table 4. Total Cost per passenger-journey comparison at 25,000 daily trips

At 25,00 daily trips

Transit mode Average cost per passenger-journey
LRT 4,13 €
Metro 4,12 €
Commuter rail 3,97 €
City Street Bus 2,53 €
AGT 2,19 €
Bus Way 2,08 €
PRT 1,19 €

17 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 8. Total Cost per passenger-journey comparison at 25,000 daily trips

Average cost per passenger journey at 25,00 daily trips

4,50 €
4,13 € 4,12 €
3,97 €
4,00 €

3,50 €

3,00 €
2,53 €
2,50 € 2,19 € 2,08 €
2,00 €

1,50 € 1,19 €
1,00 €

0,50 €

0,00 €
LRT Metro Commuter City Street AGT Bus Way PRT
rail Bus

Therefore, PRT costs per passenger trip (10 km) is only 78 % of the corresponding bus cost at low
loads, and only 47 % at 25,000 passengers per day. Compared to the other three dedicated rail modes
at a load of 25,000 passengers per day, PRT cost per passenger journey is less than a third (29 % -
30 %) of the corresponding LRT, Metro or Rail cost.

18 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

5 CONCLUSIONS AND REFLEXIONS

5.1 Conclusions
In this cost comparison study, the capital costs for the guideway and the vehicles as well as the
operating and maintenance costs have been compared for the following modes of urban transport:
• City street bus
• Bus Way (on dedicated bus-lane)
• Light Rail Transit (LRT)
• Metro
• Commuter rail
• Personal Rapid Transit (PRT)

In some cases, also Automated Guided Transit (AGT) and the private car have been included in the
comparisons and all costs are expressed in the 2002-2003 price level. Traditional transit costs reflect
the cost level in the Stockholm Metropolitan area in 2003. PRT costs are derived from several sources.
The cost model is calculated in the range from 5,000 and up to 250,000 daily trips and reflects a
typical 10 km passenger-journey. The following conclusions can be made, on the basis from the
findings of this research:
• To summarize the final findings of this research task, one might conclude that PRT is the
cheapest urban public transport mode among the studied ones, such as Bus, LRT, Metro and
Commuter rail
• Traditional public transport systems such as Bus, Light rail and Metro is about 70 % more
expensive to operate than the new and innovative transit systems such as PRT.
• A Personal Rapid Transit system is more than 40 % cheaper than traditional Light Rail
systems in operating costs.
• A Personal Rapid Transit system is three times cheaper than traditional Light Rail systems in
investment cost.
• In a sensitivity analysis, with much shorter economic life time for PRT track, vehicles and
stations, the LRT mode is still 2,6 times more expensive than PRT at the 25,000 demand level,
the PRT cost per passenger-journey will become 25 % cheaper than the BusWay, 37 %
cheaper than the Street Bus and 62 % cheaper than the LRT mode.
• Compared to the bus mode, the total producer and user cost for PRT is 55 % lower;
compared to LRT PRT is 40 % cheaper, and compared to the heavy rail modes PRT costs are
50 % and 56 % cheaper respectively. For the bus mode, user cost is 5 times higher than the
producer cost. Fore heavy modes they are fairly equal, while for LRT, PRT and AGT user
costs are between 1.6 and 3.2 times higher than the producer costs.
• PRT is the cheapest mode also when the user costs are included into the producers’ costs

In a comparative example a budget for public transport investment of 125 mUS$, or 111 m€ has been
considered. A typical LRT corridor line with a PRT loop system has then eben compared. What could
be obtained for the same amount of investment money? The conclusion is therefore, that the served

19 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

area, and, consequently, the number of served residents, would be four times higher with PRT
compared to the LRT system. Also, the door-to-door travel time might be roughly only half (or
slightly more) with PRT compared to LRT. This is not only due to the much shorter waiting times, but
also attributable to the total absence of necessary deceleration, stand still and acceleration times at all
the intermediate stops ‘en route’ with the PRT, with its off-line stations.

5.2 A personal reflexion

Another personal reflexion is the following one:
At the time when the final decision should be made as to choose a metro, rail, LRT or a bus solution,
or even a PRT-system, most decision-makers are focusing their attention to the investment costs only,
or at best, including the first year of operations. And then the conclusion often end ups at the lowest
cost alternative, which usually is the street bus (as long as only the traditional modes are considered).

However, regarding the whole economic life time of the fixed equipment, such as the guideway, say
30, 40 or 60 years (in Sweden we calculate the economic life time of the rail track to be 60 years), then
the total cost structure must consider the (real) increase in operating costs as well.

Now, for the bus mode, for example, some 60 % of the total operating costs are made up by driver
wages. Usually, at normal GDP per-capita growth rates, such as 2.5-3 per cent annually, operating
costs might augment by some 1.5 to 1.8 per annum. Over a 60 year period, this means that the
operating cost would increase by a factor of two to three times. This, in turn, will shift the relationship
between capital and operating costs, substantially, over time in favour toward more capital oriented
modes of transport. To my view, this aspect is often highly neglected.

The Figure 9 could serve as an illustrative example of this statement:

20 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 9. Cost increase of operating cost per seat-kilometre between 2000 and 2003 in real terms for
various transit modes in Stockholm, Sweden

Development of operating costs

per seat-kilometre 2000-2003 in Stockholm
Percent change in €/seat-km at fixed price level
20% 18%

15%
12%

10%

5%
5%

0%
-1%
-3%
-5%
Commuter rail Metro All Publ. Trp. LRT Bus
Modes

As can be seen from Figure 9 the Bus mode becomes much more expensive even in such a short
period of time. One reason for the negative operating cost development for commuter rail was a new
procurement in 2000. This aspect is also a strong argument for PRT as a driverless mode of transport.

21 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

6 REFERENCES
1 Cost estimates obtained from www.taxi2000.com ; www.atsltd.co.uk ; and www.austrans.com

2. Dunning, B. and Ford, I (ed): The “ATRA Report”: Personal Automated Transportation - Status and
Potential of Personal Rapid Transit; Advanced Transit Association. September 2002.

3 European Commission, DG Research, 5th Framework Programme, Key Action: “City of Tomorrow and
Cultural Heritage”, EDICT Final Report, Deliverable 10, December 2004

4 European Commission, DG Research, 5th Framework Programme, Key Action: “City of Tomorrow and
Cultural Heritage”, EDICT Huddinge Site Assessment Report, June 2004 (Ed: G Tegnér, I. Andréasson,
N.E. Selin)

5 Fabian, L.: Data-base. Trans 21

6 Lowson, M.. Service Effectiveness of PRT vs. Collective Corridor Transport. ATS Ltd, Nov 2002

7 Stockholm Chamber of Commerce: “On the Track- Prosperous Public Tansport Projects in Stockholm”.
Report from Transek Consultants. February 2005

8 Storstockholms Lokaltrafik Annual Report 2003

9 Tegnér, G. and Andréasson, I: Personal Automated Transit for Kungens Kurva, Sweden - a PRT system
evaluation within the EDICT project. 9th APM 2003 Conference, Singapore, Sept. 2003

10 Tegnér, G.: EDICT – Comparison of costs between bus, PRT, LRT and metro/rail. Transek, February
2003

11 Tegnér, G.: Ridership Analysis; in: Personal Automated Transportation- Status and Potential of Personal
Rapid Transit. The Advanced Transit Association, January 2003. (ATRA).
http://www.ianford.com/prt/ridership.pdf

12 Tegnér, G.: Comparison of Transit Modes for Kungens Kurva, Huddinge, Sweden; 8th APM Conference
in San Francisco, July 2001.

13 Tegnér, G.: Benefits and Costs of a PRT system for Stockholm. 7th APM Conference. Copenhagen, May
1999.

14 Tegnér, G., “Market Demand and Social Benefits of a PRT System: A Model Evaluation for the City of
Umeå, Sweden”. Infrastructure, Vol. 2, No. 3, pp. 27-32, 1997, John Wiley & Sons, Inc.

15 Yoder, S. “Capital Costs and Ridership Estimates of Personal Rapid Transit. Jan 2000.
http://faculty.washington.edu/jbs/itrans/yoder.htm

22 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

7 ACKNOWLEDGEMENTS

This work has been possible to carry out thanks to an original Consultancy work, commissioned to
Transek Consultants by the Stockholm Chamber of Commerce in December 2004-January 2005.

I would also like to thank Tech Dr. Jan-Erik Nowacki at Swede Track Systems AB and at the Royal
Institute of Technolgy, Stockholm, who has kindly assisted me in the early phase of constructing the
cost model.

I am also very grateful to my fellow and college, Mr. Christian Nilsson, a Transek Consultants, who
brilliantly completed the cost model work in early 2005 in close collaboration together with the author.

Finally, I would like yo thank Mr. Jan Johanson, of Ecomitech AB, who kindly assisted me in editing
this paper¨and cheched all figures.

For all potential mistakes, I am solely responsible.

23 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

APPENDIX 1: PRT COSTS

The PRT costs are divided into capital costs and operating costs.

Capital costs
For the capital costs, we start our overview with the summary cost statement made by ATRA2 a few
years ago in their excellent PRT Report3. Thereafter follows a brief summary of the costs for the three
PRT systems: Taxi2000, ULTra and Austrans.

ATRA cost estimate

Capital cost of 10-km system (system-independent)
“A customer could reasonably expect to procure PRT at the rates listed below, or could use these for
urban planning purposes. The customer could multiply the final result for a 20-km system, 30-km, etc.
For a very informal method of costing PRT. The table is listed for 10-km, which is the distance of all
the network segments added up.

Table 5. ATRA’s capital cost etstimate for PRT

Component Unit Cost Number Total (k$)

Guideway – straight 2,300 k$/km 8 18,400
Guideway – curved 3,400 k$/km 2 6,800
Vehicle 38 k$ each 100 3,800
Stations @ 2/km 250 k$ each 20 5,000
TOTAL 34,000

Therefore capital costs might be around $34 M for a 10-km network. (for preliminary planning
purposes)

That’s $3.4 M per km, or a little over $5 M per mile.”

Cited from the ATRA report op.cit.

Taxi 2000
Professor J. Edward Anderson the at Taxi 2000 has kindly provided me with detailed cost estimates
for varios PRT networks over the years. The Taxi 2000 costs were specified at a very detailed level,
not only per track, station and vehicle, but also including:
• Depot cost for 400 vehicles
• Control & communication system costs per km

2
ATRA, i.e. the Advanced Transit Association
3
PAT- Personal Automated Transportation - Status and Potential of Personal Rapid Transit. Executive Summary. January 2003

24 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

• Electric supply & distribution per km

• Maintenance, cleaning, control eqiupment, admin per km
• Engineering, assurance, marketing costs per km

ULTra

Professor Martin Lowson at ATS Ltd has kindly provided me with ULTra capital and operating cost
data. The capital costs are given as a formula that depends on:
• A fixed cost per track.-km
• The rate of elevation
• The number of passengers per year (utilization rate)

Cost calculations have been made for many potential sites such as Cardiff in Wales, King’s Curve,
Huddinge in Sweden and several sites in United Kingdom.

Exerpt from the EDICT-report about ULTra cost in Cardiff:

“At this early stage of the project, estimates of the costs of the system are unavoidably preliminary,
and will require more detailed engineering studies to refine them. Nevertheless, the mean costs of the
infrastructure are based on detailed engineering estimates, and confidence in them is engendered by
the fact that construction of the test track was achieved within these costs. Moreover, two experienced
construction contractors have estimated the costs of constructing the infrastructure for a larger 19.8km
network extending beyond the case study network. One contractor estimated £33.2M, the other
£43.9M, compared with the ULTra Consultant’s estimate of £67.9M. The Contractors’ estimates do
not include the cost of mechanical and electrical equipment and track electrical/ control installations,
which add about 13 % of the cost, whereas the Consultant’s costs include this, and there was an
increase in steel prices between the lower and higher Contractor’s estimates. The Consultant’s
estimate is very conservative because of concerns that building above existing streets might increase
costs, but the costing used in this assessment adds a further 15 % contingency to the Consultant’s
estimates, and in relation to the Contractors’ estimates they offer a total contingency margin of some
70 %. Thus even though the scheme is innovative, the costing seem very robust indeed.”4

This means that instead of the calculated £67.9M (or 5.7 M€ per track-km) the Contractors’ bid price
is around 3.7 M€ per track-km, or 36 % lower.

4
Source: EDICT, European Commission: “Cardiff Site Assessment Report. Deliverable 6-2. Version 1.5. April
2004. Authors: Advanced Transport Systems (ATS), Ove Arup and Partners International Ltd (ARUP), The
County Council of the City and County of Cardiff (CCC), Transport & Travel Research Ltd (TTR)

25 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

AUSTRANS

Form Director Phillip Robinson of Austrans, I received in 2003 some cost estimates for Austrans’
potential application sin Singapore and in Leipzig, Germany. On example is presented below for a 20
km GRT track:
Table 6. Austrans GRT cost estimate

Austrans GRT 20 km track, 40 stations

m€/km
Investment cost 9 per track.km
Operating cost 0,08 per pass.km

PRT Capital cost comparison

An attempt to compare the three PRT suppliers’ cost estimates is presented in Table 7 below
Table 7. Authors’ own cost comparison of PRT capital costs based on suppliers’ data

PRT system Taxi 2000 ULTra Austrans ATRA average # of units

Unit cost m$ m$ m$ m$ for a 10 km track
track per km 1,5-4,2 1,5 5,4 2,75 10
per Station 0,26 2,7 incl other 1,2 0,25 20
per vehicle 0,03 0,05 0,1 0,038 100
Other costs included included included included
Total cost for a
10 km PRT system, m $ 23,1-53,6 51,4 91 34

From Taxi 2000 cost estimates have varied a bit over the past 5 year period. The comparison is
presented for a 10 km PRT track. To achieve the same passenger level-of-service, I have assumed the
same amount of vehicles and stations for the three systems. Austrans’ vehicle carries 9 seated
passengers, while ULTra carries 4 and Taxi 2000 3 passengers. Austrans vehicle fleet thus gets an
over-capacity compared to the two other suppliers.

The total PRT capital cost varies between 23 and 51 M $ per 10-km track. The Ultra application for
King’s Curve in Sweden amounts to 69 M $ (with 100 % elevation). This can be compared with the
ATRA average for a true PRT system amounts 34 M $.

One might conclude that the capital cost for a PRT network is quite uncertain as regards its magnitude,
as there is not any real market for mass production of PRT-system, yet. However, most of the careful
concept studies, that have been carried out so far, indicate a cost range from 2.3 to 9.1 M $ per track-
km, with an average of 5,8 M $/km. In the cost model I have used 6 M$/track-km.

26 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

PRT Operating Costs

The operating cost experiences stem from the Greater Stockholm Transit Company; AB Stor-
Stockholms Lokaltrafik, Annual Report 2001. The PRT Operating costs have been obtained from
each supplier in January-February 2003. The American experiences are from the Trans 21 Database
(Lawrence Fabian’s database). The following results that were found also include some other modes
of transport, in order to broaden the comparison:
Figure 10. Operating & maintenance cost per passenger-km comparisons, 10 cases

O & M Cost/passenger-km in Euro

12 LRT-systems in USA 0,21

Suburban Train+Tramway-Sthlm 0,21
LRT-Sthlm: Gullmarsplan-Alvik 0,21
Bus-Stockholm 0,18
PRT - ULTra 0,16
Metro-Stockholm 0,16
22 Automated Guided Systems in USA 0,12
Commuter rail-Stockholm 0,10
PRT - Austrans 0,08
PRT - Taxi 2000 0,06

0,00 0,05 0,10 0,15 0,20 0,25

As can be seen from the figure above, PRT systems perform very well also from the operating cost
point-of-view. Most LRT-systems, both in the US and in Stockholm, cost around 0.20 Euro per
passenger-kilometre and the bus network (9 153 km; 1 674 buses) in Stockholm costs 0.18
Euro/passkm.

The ULTra PRT cost figure (calculated for the first phase network at Kungens Kurva) show a lower
cost, 0.16 Euro per passenger-km, which is the same level as for the Stockholm Metro system (108
km; 800 metro-cars). The average operating costs for the 22 Automated Guided systems in USA cost
0.12 Euro per passenger-km and the Commuter rail network in Stockholm (186 km: 292 vehicles) cost
0.10 €/passkm.

Both the Austrans and the Taxi 200 PRT systems are even cheaper in operation with 0.08 and 0.06
Euro per passenger-kilometre respectively.

An average of the three PRT systems yields an operating (and maintenance) cost of 0.10 Euro per
passenger-kilometre. The average operating & maintenance costs for all the other systems (AGT/LRT;
Bus, Metro and Commuter rail systems) are 70 % more expensive to operate than for the PRT system.

27 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

The Operating & maintenance costs comparison is summarised in the figure below:
Figure 11. Operating & maintenance cost per passenger-km comparisons, 2 cases

O & M Cost in Euro/passenger-km

0,18 0,17
0,16
0,14
0,12
0,10
0,1
0,08
0,06
0,04
0,02
0
39 LRT/AGT or Bus systems 3 PRT systems

28 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 12. Comparison of Operating Costs in Stockholm – traditional modes vs. PRT

Comparison of Operating costs per trip in Stockholm: PRT and

traditional public transport modes

£2,00 1,89 €
1,79 €
£1,80

£1,60

£1,40

£1,20 1,15 € 1,14 €

£1,00 0,93 €
0,79 €
£0,80

£0,60

£0,40

£0,20

£0,00
PRT-Kungens Average trad. Metro Bus LRT Commuter rail
Kurva PT in
Stockholm

In Stockholm, only the commuter rail system has such low operating costs as PRT. This is presented
in Figure 12 above.

29 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

APPENDIX 2. COST COMPARISON

Capital costs for guideway

The investment or capital costs for traditional modes are divided into capital costs for the fixed
equipment, i.e. the track and stations and for the rolling stock, i.e. the vehicles.

The capital costs for the infrastructure are derived from recent new investment schemes for LRT,
Metro and Commuter rail projects in the Metropolitan Stockholm Area.

Figure 13. Capital cost per track-kilometre in million Euro

Capital cost (excl. vehicles) per track km, M€

Traditonal public transport = Stockholm data
120
111,1

100

80

60
46,0
40
28,0

20 15,8 15,4
4,0 3,0
0
City Highway Metro Commuter 22 AGT LRT 3 PRT BusWay
(tunnel) rail systems systems Amsterdam

In Figure 13 the capital costs per track-km for the three Stockholm rail modes: LRT, Metro and
Commuter rail are compared also with:
• 22 International AGT systems (AGT= Automated Guided Systems)
• 3 PRT systems (Taxi2000, ULTra and Austrans)
• A separate Busway in Amsterdam (the Zuidtangent bus corridor)

The average PRT track cost amounts 4 M€/km excluding vehicles, which is roughly in the same
magnitude as for the Busway in Amsterdam; but only one-fourth of the Stockholm LRT- project
“Tvärbanan”, phases 1 and 2, now in operation. The capital cost for rail is more than 7 times higher
than for PRT, and the metro track cost (including stations) is eleven times higher than for PRT per
kilometre. A city highway in tunnel (as the Southern Link) opened last year in Stockholm, costs more

30 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

than 110 million Euro per kilometre, or 28 times higher than PRT. (Of course there is also a difference
in passenger capacity, see section 4.1 below).

Capital costs for vehicles

Following the assumptions above we end up with the results presented in Table 8 below:
Table 8. Vehicle capital costs and capital cost per passenger-kilometre for seven modes of transport

Com-muter
LRT Metro car rail Passen-
Key indicators Citybus (2 cars) 2000 ("Regina") AGT PRT ger car
Capital cost/vehicle, m€ 0,31 2,22 3,33 5,00 2,78 0,05 0,02
Economic life time, years 18 25 25 25 25 8 10
Discount rate 4% 4% 4% 4% 4% 4% 4%
Annuity factor 0,078 0,063 0,063 0,063 0,063 0,146 0,121
Annual captal cost, m€ 0,024 0,141 0,211 0,316 0,176 0,007 0,003
Operating hours/year 3000 3500 3700 4000 5600 7000 525
Days/year 350 350 350 350 350 350 350
Operating hours/day 9 10 11 11 16 20 1,5
Average speed, km/h 15 25 44 50 50 36 36
Vehicle-kms/year 0,045 0,088 0,163 0,200 0,280 0,071 0,019
Capital cost/vehicle-km, cent 53 161 130 158 63 10 14
Seats/vehicle 46 78 126 100 78 4 5
Capital cost/set-km, cent 1,15 2,06 1,03 1,58 0,80 2,57 2,85
Load factor: person/seat 0,269 0,232 0,366 0,279 0,366 0,370 0,320
Capital cost/pass.km, cent 4,28 8,86 2,81 5,66 2,20 6,94 8,92

The operating figures (vehicle-hours and –kilometres, seating capacity) and the vehicle capital costs
for the traditional modes are derived from Stockholm experiences. The AGT capital vehicle cost data
was obtained from Lawrence Fabian’s (at Trans21) Data Base. The capital vehicle cost per passenger-
kilometre is presented in Figure 14 below:

31 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 14. Capital vehicle cost per passenger-kilometre

Capital vehicle cost in Euro-Cents/passenger-km

10,0
8,9 8,9
9,0

8,0

7,0

6,0 5,7

5,0
4,3
4,0 3,7
2,8
3,0
2,2
2,0

1,0

0,0
Passenger LRT Commuter Citybus PRT Metro car AGT
car rail 2000

The private car and the Light Rail Transit (LRT) vehicle costs are of equal magnitude, around 9 cents
per passenger-kilometre. Commuter rail vehicle cost is about 6 cents, and the metro less than 3 cents.
The City bus and the PRT vehicle have roughly the same size of the capital vehicle cost, around 4
cents. The AGT vehicle comes out with about half that amount, 2 cents. The metro car seems to get a
lower vehicle cost per passenger-kilometre (around 3 cents/pass.km), probably due its very high
passenger-loads.

In this respect, vehicle capital cost per passenger-kilometre, PRT does not get the lowest cost. The
corresponding PRT costs will be more elaborated in chapter0 above.

Economic life time and discount rate

In order to be able to compare investment oriented transport systems, as heavy rail, with operating
cost-oriented systems as the urban bus in a common framework, we consider the economic life times
of each system as well as the discount rate.

In Sweden a common practice for social cost-benefit analyses is to use a 4 % discount rate.
The economic life times are assumed to the following ones:
Figure 15. Assumed economic life times for various modes of transport- track and vehicles

32 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Capital cost (excl. Vehicles)/track km, M€

Traditonal public transport = Stockholm data
120
111

100

80

60
46
40
28

20 16 15
4 3
0
City Highway Metro Commuter 22 AGT LRT 3 PRT BusWay
(tunnel) rail systems systems Amsterdam

The economic life time of the various vehicle types can always be discussed and questioned. Maybe
the automatic systems, such as AGT and PRT, will have shorter life times in reality, than has assumed
here. The corresponding PRT costs will be more elaborated in chapter 0 below. In section 4.6 we
present some sensitivity tests

Operating costs

The operating costs are dependent on a broad variety of factors, such as the level-of service, labour
unit costs, efficiency and the technology of the public transport mode in question. For the traditional
public transport modes, Stockholm data from 2002 has been used. The BusWay operating cost is
calculated to be 28 per cent lower compared to the city bus due to its higher speed on its own, separate
right-of-way. The AGT operating cost was obtained from Lawrence Fabian’s (at Trans21) Data Base.

The PRT operating costs are obviously a bit more uncertain, due to the lack of real life experiences
from PRT operations. This estimate is based on cost judgements made by the three potential suppliers,
Taxi2000, ULTra and Austrans.

33 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Figure 16. Operating and maintenance costs per passenger-kilometre

O & M Cost, €/passenger-km

0,30

0,25
0,25
0,21
0,20
0,16
0,15
0,15
0,12
0,11
0,10
0,07

0,05

0,00
LRT City Bus Metro BusWay AGT Commuter PRT
rail

Light Rail Transit - indeed not very light, but a rather heavy mode of public transport – is the most
expensive system in terms of O &M costs per passenger-kilometre. The difference to the second most
expensive mode to operate, the City bus, is not that big, 0,25 € and 0,21 € per passenger-km,
respectively. The BusWay (like the Zuidtangent BusWay in Amstersdam) has a similar cost level for
operations comparable to the Metro, i.e. rather cost efficient with its 0,15-0,16 € per pass.km.
Commuter rail is even more cost-efficient, mostly due to long train lengths with many cars in the
trains, and very few crew members per passenger-journey.

AGT and Commuter rail has less than half the operating cost level per passenger-kilometre compared
to the LRT system, 0,11 compared to 0,25 €/pass.km

PRT is even more cost-efficient, with its estimated 0,07 € in operating cost per pass.km. This is less
than a third of the O & M cost for LRT.

34 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

APPENDIX 3. OTHER COMPARATIVE COST DATA

Transek’s investment cost comparison

The total investment costs for guideway, vehicles and stations are compared to ten different systems:

Investment Cost in M Euro per track-km

LRT Line 2 in Montpellier, 19,5 km 21,8

LRT Syd Flemingsberg-Älvsjö 20,7
LRT in Orleans 19,8
LRT Sthlm:Alvik-Gullmarsplan 19,3
22 st Automated Guided Systems in USA 17,2
LRT in Barcelona 16,0
Rubber-tyre LRT, Caen 14,4
LRT Sthlm: Gullmarsplan-Sickla 14,4
PRT - Austrans 8,0
Phileas Bus in Eindhoven 8,0
Trolley bus in Orléans 7,2
Busway in Amsterdam -Zuidtangent 6,0
PRT - ULTra 5,9
PRT - Taxi 2000 3,4

0,0 5,0 10,0 15,0 20,0 25,0

Bars in blue colour are Stockholm systems, in green are PRT systems and in brown are other systems.
The Three Stockholm LRT system costs around 15-20 M Euro per track-kilometre5. The new LRT
Line 2 in Montpellier costs approx. 22 M Euro per km. A summary of 22 Automated Guided systems
in the US have an average cost of 17M Euro per track-km6.

All three PRT systems show a lower investment cost than the studied LRT systems. With Austrans at
9 M€, ULTra at 5,6 M€ and Taxi 2000 at 3.4 M€ per kilometre7. An average of the three PRT
systems yields an investment cost of 6M€ per track-kilometre8. This is the same cost as for the
newly opened Zuidtangent Busway outside Amsterdam and even lower than the Trolley bus in
Orléans, France.

The average investment costs for all the other systems (AGT, LRT and Busway/Trolleybus route) are
more than three (3!) times higher than for the PRT system.

5
Source: AB Storstockholms Lokaltrafik
6
Source: Data-base collected by Lawrence Fabian, Trans21.
7
Sources: Cost estimates from the Suppliers
8
We used this average value in our benefit-cost analysis for Kungens Kurva in the EDICT-project.

35 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

Conclusion: A Personal Rapid Transit system is three times cheaper than traditional Light Rail
systems in investment cost.

A simplified comparison between 29 various LRT/AGT or Bus systems (se page above) with the 3
PRT systems, show the following investment cost differences:

Investment cost per track-km in M Euro

25

19,7
20

15

10

6,0
5

0
29 LRT/AGT or Bus systems 3 PRT systems

PRT is therefore a low-cost system from the investors pint-of-view.

The Orléans Cost study

In the Orléans Cost Study the following costs were identified for the modes: Bus, Trolley Bus, Rail
Bus, Rubber tyre LRT and Tramway:

Cost Study Orléans Key facts

Source: Rail & Transport, 3rd July 2002
Line length in km 22,7
Stops 34
Vehicles 25
m€
TOTAL COST various modes m Euro Investment cost/km
Tramway 321 14,1
Rubber tyre LRT 290 12,8
Rail Bus 175 7,7
Trolley Bus 165 7,3
Bus 125 5,5

This Study reveals that Rail and Trolley bus cost about half as much as Tramway in investment per
kilometre and that “Bus” cost half as much as Rubber tyre LRT. The cost level can be compared with
the figures presented in Figure 15 above.

36 Transek AB
PRT Cost Comparison – AATS Conf 7-8 Nov, 2005

The PHILEAS BusWay in Eindhoven

The Phileas Automated Guided Bus, (AGB) in Eindhoven consists of a 15 km BusWay with 20 stops
and 12 buses. It has a capacity that ranges from 48 to 24 passengers. The investment costs were:
• Infrastructure: 70 M€
• Vehicles: 40 M€
• Control system: 10 M€

The total capital costs were therefore 120 M€, or 8 M€ per Busway kilometre. This is mote than the
estimated PRT Capital cost at 6 M/track –kilometre.

Capital Costs and Ridership Estimates of Personal Rapid Transit

In an in-depth study on PRT Costs, Supin Yoder at Wilbur Smith, presented a study on Capital costs
and Ridership of PRT for the projected PRT system in Rosemont, Chicago in 2000.

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 were examined – (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.

Three analysis techniques were used – (1) statistically significant regression analysis; (2) measurement
of a central tendency for "comparable" systems; and (3) statistics from all AGT systems. The results
show that the combination of the three techniques worked well for component-level studies and show
promise for use in other cost analyses involving new technologies or application of existing
technologies on a scale outside the bounds of previous experience. In addition to the cost study, a PRT
ridership forecasting approach and projections were evaluated, providing another key element of
decision support for potential PRT deployment in Rosemont.

Her conclusion were:

• The PRT-comparable cost range tends to be on the lower side of the cost ranges for the 17
AGT systems.
• Raytheon’s cost projections for five of the seven components are comfortably in the
comparable component cost range.
• From the system cost perspective, Raytheon’s projection is in the range of PRT-comparable
AGT system cost. If initial PRT deployment or nonrecurring costs are excluded, Raytheon’s
PRT unit cost projection is significantly lower than the comparable AGT unit cost.

Even this Cost study indicates that PRT capital costs are lower than the corresponding AGT costs.

37 Transek AB