Paper8 by hcj


									Paper 8

Author: Göran Tegnér, Transek AB, Sweden

Title: The use of EMME/2 for an improved Bus Network in the city of Örebro, Sweden


    0   ABSTRACT.................................................................................................................................... 1
    1   The City of Örebro and its Bus Network......................................................................................... 2
    2   Today’s Public Transport and Road Traffic Flows.......................................................................... 2
    3   The demand model and coding the network .................................................................................. 4
    4   Attracting new riders by improving the bus network? ..................................................................... 7
    5   Six bus network scenarios analysed .............................................................................................. 9
    6   Trip characteristics – link flows ...................................................................................................... 9
    7   Overall results and recommendations.......................................................................................... 12


The bus network in the central parts of the city of Örebro, a medium sized city in mid-Sweden, has been
analysed by applying the FREDRIK Demand model and the EMME/2 Transportation Planning System.
The city of Örebro has a dense and appropriate bus network. Nevertheless, there is a scope for
improvements in the network.

In the study we have used various modules and options in the EMME/2 model, in order to analyse the
market for improved transit ridership. Most transit companies undertake several on-board surveys, and
extract information about the present day trip pattern – among already captured transit users. However,
to get new riders, there is a need to know the trip pattern of the non-users, mostly car users. By
calculating the original and destination pattern for the car trips, we have revealed the trip pattern for the
competitive mode – car users. We also created a transit travel time elasticity matrix, to find out where
the highest responses to transit travel time reductions were to be found. This was all done with the

In the city centre many bus lines are concentrated. Most plot pictures of transit flows thus become rather
messy in a traditional EMME/2 plot. By using the new feature with extra link attribute in module 2.13 in
release 9.0, we have been able to colour separate the various transit lines according to different types,
such as: city buses, county buses, and regional rail lines. A comprehensive study of six network
scenarios were then carried out with the following type of output:
          Transit passenger flows by line & car traffic volumes
          Overall mode shares transit – car traffic
          Travel time gains by travel component
          The Generalised Cost concept (weighted travel times and costs)
          Number of boardings, alightings and transfers at nodes
          Total vehicle kilometres, vehicle hours and average vehicle speed

Our findings indicate that a concentration and a modification of four major bus lines into a so called:
“Head lines” in a combination with altered itineraries for the county bus lines from more distant
communities via suburban work-places would enhance the use of public transport and reduce the
overall generalised travel times for the users. This would also marginally reduce car traffic in the central
parts of Örebro.

1       The City of Örebro and its Bus Network
The city of Örebro, 150 km west of Stockholm has population of 72.000 inhabitants 7while the entire
county of he Örebro contains 233.000 people. The number of employed are 33.000 persons in the city
and 113.000 persons in the county.

There is an inter-city rail service as well as a new high-speed regional commuter rail service between
Stockholm and Örebro. There is a large regional hospital and a University in Örebro, which is also a
regional centre for industry and Örebro is the regional administrative hub of mid-Sweden.

2        Today’s Public Transport and Road Traffic Flows
The public transport system is dominated by local and regional bus service, and the commuter rail
service, which is served by railway stations in the city of Örebro (2 stations). The structure of the bus
network is completely radial with all lines focused in the central squares in the city centre.

The map shows the central area of Örebro covered by the FREDRIK - EMME/2 Modelsystem:
The blue lines show which roads/streets are geo-coded into EMME/2, while the full underlying road network is
presented as a MapInfo background layer.

The map below shows the regional area covered by the FREDRIK - EMME/2 Model system.
Not every road and regional bus line is coded into the model. But all major primary and secondary roads
and all bus lines that are connected to the central parts of the city of Örebro are included.

3       The demand model and coding the network

The FREDRIK – EMME/2 Model system
TRANSEK Consultants Company has developed the traffic planning system FREDRIK, which is a com-
puter model system to handle travel demand in urban and regional areas in a user-friendly way. This
planning tool consists of a system of logit models simulating travel behaviour in terms of choices of the
trip-makers. This comprises. the number of trips generated, between which areas, by which mode of
transport the trips are made and along which travel routes and transit lines the trips finally are decided
to take place. This system is integrated with the world famous EMME/2 transportation planning
package. Together the two systems - FREDRIK and EMME/2 form a complete four-step transportation
model system for integrated road and public transport planning for urban and regional applications.

                 FREDRIK - EMME/2
  The 4-step Travel Demand & Supply Model System

      Future Transport
                                 Land use and
      Network                                                     Other Traffic
      * Auto                                                        variables
      * Transit

      EMME/2 Network
      Assignment Model

      Trip Travel Times
      * Auto
      * Transit


      External flows from
        regional entries       Auto and Transit        Calibrated Auto
                                Trip Matrices         and Transit flows

                             EMME/2 Network
                             Assignment Model

                               Assigned and
                              Calibrated Trips

               system consists of eight sub-models for each of the
The FREDRIK and stands for Frequency choice,Route choice, Emme/2, following trip purposes:
      FREDRIK is a registred trade mark of TRANSEK Consultants
 Work-trips
      Destination choice, Route mode choice, logIt models, transeK.

 Home-based Business trips
 Work-place-based Business trips
 Shopping trips

 Personal Business trips
 Recreation trips
 Personal Visit trips
 School trips
The mechanism of the system implies that the number of trips generated, the distribution of trips on
destinations as well as the choice of travel modes, are highly dependant on a wide range of explanatory
factors, all of them treated in a behaviourally sound and consistent way. The system handles four travel

 car transport
 public transport

 walk mode
 bicycle mode
With the FREDRIK/EMME/2- model system the following impacts can be analyzed:

 Road Network
    - new road/street links
    - closure of road capacity (bus-only
      lanes, one-way street)
    - changed road/street capacity
    - road pricing

 Public Transport Network
    - new lines (bus, train)
    - changed level-of-service (times,
      speeds, frequencies, transfer points,
       walking distances)
    - public transport fares

 Residential and Work-place Location
    - new residential areas
    - new work-place areas
    - changes densities/urban sprawl/
    - changed pattern of activities

 Travel Costs
     - costs for auto usage
     - costs for transit usage

 Socio-demographic factors
    - household size and income levels
    - car ownership levels
The FREDRIK system is used to simulate various types of transportation policies. The impacts from all
kind of changes on the total amount of road traffic will be calculated by using the EMME/2 equilibrium
method to achieve equilibrium flows in the road network, using volume-delay functions for simulating
congestion. Also for the transit mode, the EMME/2 optimal strategy method is used. The impacts on
the car and transit trip patterns from altered land-use, as well as from changes in the socio-demo-
graphic structure of the travellers could also easily be analysed within the framework of the model
system. This four-step model is in use (May-99) in Stockholm and the Mälar Valley Region as well as
in Malmö, Uppsala, Västerås, Örebro, Umeå, Södertälje and Halmstad in Sweden. It is also in use in
Oslo, Norway. The system is installed in all Road and Rail Administration Regions in Sweden and at
SIKA – the Swedish Institute for Communication Analyses. The system has moreover been utilized for
studies on light rail projects in the city of Kayseri in Turkey.
Network characteristics
The EMME/2 databank consists of:
          248 traffic zones
          1,096 nodes
          3,636 directed links
          100 transit lines (for three specific time-periods)
          4,000 transit lone segments
          340 turns
The planning problems
The detailed planning questions that were supposed to be answered by the study were:
     The possibilities to reach the new “travel centre” at the railway station
     The impacts of a new bridge across Svartån (the Black Stream)
     The location of bus routes within the city centre
     Improved accessibility to the Regional Hospital of Örebro (RSÖ), and to the university and
         other work-places outside the city centre
     Is there enough demand for tangential bus routes?
     The co-ordination between regional and city bus lines

4       Attracting new riders by improving the bus network?
The market share for bus trips is estimated to be 25 percent of all motorised trips (car and public
transport trips, walk and bike trips excluded). This is an extremely high market share. How do you
achieve to improve such a good system in a medium sized city like Örebro?

Well we started our analysis by comparing the origin-destination trip pattern between car users and
bus/rail users.

OD-trip pattern among car users

In the EMEM/2 plot picture below the trip destinations for car trips is shown for the south-western part
of the central area of Örebro:

As can be seen from the plot picture there is a substantial spread of trip ends among car users. Most of
these work-places are industrial zones, with extremely low market shares for public transport mode –

OD trip pattern among bus users

This trip destination plot picture shows the trip ends among bus users aiming at the city centre area:

As can be seen, the trip ends are rather concentrated to a few centrally located squares in the city of

As a matter of we found out from the trip matrices a dramatic difference I the overall trip pattern
between car users and bus users:
        The bus mode has its highest mode shares only to the central city destinations, with very
           few exceptions; and extremely low modes shares to all other - more peripheral
        The car mode has rather low mode shares to the city destinations, but on the other hand,
           very high mode shares to all other, non-central locations.

So, if one wants to improve the market for urban bus traffic, one has to seek the solutions among those
origin-destination pairs, where the market potential (low transit market shares, but very high absolute
concentrations in terms of the total number of commuters) is substantial.
5         Six bus network scenarios analysed
At first a number of bus planning ideas were tested before we introduced the network model system.
Then six different scenarios were established and compared to the base scenario:

     New bridge                Re-routing of bus lines to a new bus only bridge
     Pendulum                  Development of today’s radial bus network
     Major Bus Routes                  Concentration on a few major heavy bus lines
     Tangential                Establishing a few ring lines
     Improved Regional Bus             Give the regional bus lines a "city role"
     Double bus frequency Sensitivity analysis of a doubled bus frequency1
Performance indicators
The following performance indicators were chosen as guidelines for evaluation for the City and County
of Örebro:

 Transit market shares
 Weighted travel time gains (Generalised time)
 Increased transit ridership (absolute and in percent)
 Impact on car usage
 Supply indicator: (vehicle fleet; peak and off-peak headway; vehicle kilometres)
 Social net benefits

6     Trip characteristics – link flows
The amount of car and transit trips is shown for the central part of Örebro below:

Transit trips are much fewer and more concentrated to the city centre than car trips are:

    The doubled frequency is applied on the four heaviest bus lines- not on all lines.

In the city centre many bus lines are concentrated. Most plot pictures of transit flows thus become
rather messy in a traditional EMME/2 plot. By using the new feature with extra link attribute in module

2.13 in release 9.0, we have been able to colour separate the various transit lines according to different
types, such as: city buses, county buses, and regional rail lines.

On the regional scale, rail lines are shown in blue, regional bus routes in red and city bus lines in red.
The same colours are representing rail, regional and urban bus lines for the central part of Örebro

                                                      KOLLEKTIVA RESOR
                                                          DYGNST RAFIK

                                                                               T ag

                                                                         SCALE:         250


            EMME/2 PROJECT : rebro 250zones Mbasres                      00-02-07 13:47
            SCENARIO 14:     Dygn orebro 991022jh                        MODULE: 2.13
                                                                         T RANSEK5....jh

Already with this plot, the reader can see how regional bus lines are interfered with the local city bus
lines. A better co-ordination between these two separate sub-systems is a natural way of improving the
overall performance.

As an example from one of the analysed six scenarios, we show the difference plot for the “Improved
Regional bus” scenario compared to the base scenario:

                                                              TRANSIT VOLUMES
                                                     DIFFERENCE   SCENARIO   54 - SCENARIO   14
                                                                                                                    54 - 14

                                                                                                                  SCENARIO 54
                                                                                                                  TRANSIT LINES:
                                                                                                                    ALL LINES
                                                                                                                  SCENARIO 14
                                                                                                                  TRANSIT LINES:
                                                                                                                    ALL LINES

                                                                                                                  SCALE:         250


             EMME/2 PROJECT: rebro 250zones Mbasres                                                               00-02-04 15:57
             SCENARIO 54:    cpsc91 dygn med nya lansbussar                                                       MODULE: 6.23
             SCENARIO 14:    Dygn orebro 991022jh                                                                 TRANSEK5.. ..jh

A new itinerary for the regional buses entering the city of Örebro from the south-western suburbs and
peripheral cities in the county was proposed. Instead of entering the city via the E20 motorway, these
bus routes have been diverted through some industrial areas. The decline in the flows (in red) for the
present day itinerary is more than offset by the new “Improved Regional Bus” network (in green).
These bus lines are also prolonged till RSÖ in the north-eastern part of the inner city (the Regional
Hospital of Örebro). The number of transfers in the city centre would be reduced and some car trips
would be attracted to the transit mode instead.

7         Overall results and recommendations

Major results

Some major results are summarised in the table below:

Analysed scenarios                       Generalised                  City market Increase                 in Decrease in Vehicle fleet
                                         time                         share2      transit                     car traffic in       peak
                                                                                  ridership                               hours
Base network                             90,7 min                     25,5 %                      -           -           65
New bridge                               89,8 min                     25,6 %                      +1,9 %      -0,1 %      66
“Pendulum”                               90,4 min                     25,9 %                      +2,2 %      +0,1 %      61
Major Bus Routes                         88,4 min                     26,0 %                      +2,2 %      -0,1 %      61
Tangential                               89,5 min                     26,0 %                      +2,0 %      +0,1 %      60
Improved Regional Bus                    85,9 min                     26,3 %                      +5,9 %      -0,4 %      65
Double bus frequency                     83,2 min                     26,4 %                      +3,8 %      -0,2 %      107

A good Public Transport Bus Network could be further improved:

    Of Transit + Car traffic only( walk and bike modes are excluded)

     In the city of Örebro the motorised market share is 26 % and twice as high as for the county of
     With a transit market share of 13 % for the entire county there is a theoretical potential of 87
      % trip makers that do not go by transit today. If one might be able to recruit only 5 % of these
      87 % non-users to switch to bus or train, this would imply 33 % more transit users..
     If the Örebro Transit system should be developed so as to compete with the private car, then
      more regional bus lines should pass the peripheral induistrial areas in south-west at peak
      hours. Besides, as many regional bus lines should be terminated towards the regional hospital
      instead of just in the city centre.

    “The Major Bus Route” and the “Improved Regional Bus” alternatives are the best solutions as
     the improvement is judged to be possible to implement at the present day resource level (by
     slightly reducing the frequency of some other minor bus lines. Such an improved bus network
     might contribute to a reduction of the amount of car traffic by some 1,250 car trips per day
     (approx. 0,5 %).
    The perceived travel time gain for the two best scenarios – “The Major Bus Route” and the
     “Improved Regional Bus” alternatives is estimated to be 2.3 and 4.8 minutes respectively, or
     by 3 % and 5 % respectively, compared to the present day situation. The number of bus trips
     is estimated to increase by some 3,000 trips per day or by 8 %.

It is not sufficient only to alter the bus routes marginally. There is a need for broad and co-ordinated
package of measures in order to improve transit ridership substantially.

Such ingredients of a co-ordinated package for a major increase in transit market shares might be:

Enhanced plainness and transparency
 Information system and directed market campaigns
 A major bus network with a few major bus lines with fixed time intervals
 A conceptual network
 Optimised pricing structure

A clear structure for the bus routes
 Major bus routes
 Integrated regional bus lines in the city

Improved infrastructure
 A new bus bridge across the river
 Bus streets
 Bus priority measures

The study was well received in the City and County of Örebro. The work will now continue according to
our recommendations and an ”improved” bus network will be implemented after some more detailed


To top