Emme/2 as a Public Transport System Design Tool

1
Emme/2 as a Public Transport System Design Tool

• Meeting some South African Challenges

2
The Process in Perspective

• The Emme/2 modelling approach presented here is
  the first, and potentially standalone phase, of a
  three step public transport modelling process
  incorporating
• Integrated, multi-modal network modelling of all
  transport demand.
• Empirical conversion of network model output to
  route definitions.
• Multi-objective allocation of resources to public
  transport design.

3
The Objective.

• A practical design tool for guiding South African
  government transport policies with respect to
• Integrated metropolitan transport plans.
• Promotion of public transport.
• Tendered public transport services.
• Empowerment of small operators.
• Rationalisation Plans.
• Operating Licence Strategies.

4
Underlying Design Principles.

• Public transport must be planned as an integral
  part of the overall transport system.
• Route design must approximate private mode travel
  patterns.
• Public transport system design must be mode
  independent.
• Resources must be efficiently used.
• The most efficient system is not necessarily the
  most efficient application of any one system
  component.

5
(No Transcript)
6
Route Design Considerations.

• Routes must satisfy demand.
• Routes must follow the paths chosen by customers.
• Service frequency should be matched across entire
  system and be attractive to customers.
• Routes should be self-sustaining.
• Technology used should match demand.

7
Preparing the Network Model.

• Obtain public transport demand matrices.
• Temporal matrices with Growth factor the public
  transport design cycle is one week, not the
  morning peak period.
• Establish a design value for transit vehicle
  occupancy.
• Set policy attributes for all links.
• Set volume delay functions for ALL links on
  network.
• Modify mode definitions to make transit the
  primary mode.

8
Demand Matrices

• Public transport demand matrices, based on
  household surveys and allowing for some marginal
  growth in demand according to zone attributes
• Separate out potential express demand based on
  minimum O-D demand levels and distance
• Potential Express demand
• Regular demand

9
Vehicle Occupancy.

• Establish the capacity of one passenger car
  equivalent, road-based public transport vehicle.
• One 60-seat bus ? 4 passenger car units.
• ?Vehicle occupancy ? 60/4 15.
• One 14 seat taxi ? 1 passenger car unit.
• ?Vehicle occupancy ? 14.
• Use 14 passengers as average vehicle occupancy
  rate.

10
Policy Attributes.

• Mark network links according to a policy
  structure. For example
• Promote public transport on some links.
• Discourage public transport on some links.
• Links used in spite of discouraging policy may
  require functional re-evaluation.
• Never disallow public transport on any link.

11
Volume Delay Functions.

• All links carrying public transport must have
  volume delay functions.
• For road and pedestrian modes, these will be
  conventional functions.
• Non-road mode functions must account for
• Vehicle occupancy global occupancy rate not
  applicable.
• System capacities travel time independent of
  demand.
• Technical factors minimum headway.

12
The Network Modelling Process.

• Generalised-cost, multi-modal, multi-class
  assignment of all demand on network with transit
  demand as primary class.
• Extraction of travel times from multi-class
  assignment.
• Extraction of potential express demand from
  demand matrices.
• Iterative generalised-cost, multi-modal, single
  class assignment to focus public transport travel
  paths.

13
Generalised-cost Function for Public Transport.

• Users perceive a cost of travel on each link as
  a function of a number of factors
• Travel time perceived equally by all users.
• Link length / fare component introduces modal
  cost differences.
• Demand penalty encourages users to travel in
  groups.
• Policy factor promotes or discourages use of
  certain links.

14
Demand Penalties.

• Demand penalties applied to motorised links only.
• Demand penalty increases with
• Decreasing two-way link volume.
• Decreasing directional balance.
• Increasing disparity in demand between operating
  periods.
• Determined using a Link Index.

15
The Link Index.
The general form for the link index LI.
16
Determining the Demand Penalty.

• For a single operating period model
• Demand Penalty DP (MF / LI) 0.9
• For a multiple operating period model
• Demand Penalty DP (MF / CLI)) 0.9

17
The Focusing Process.

• Determine link indices for each motorised link in
  each operating period.
• Determine the composite link index for each link.
• Determine the demand penalty for each link.
• Determine the new fixed cost.
• Weight of fixed cost 0.1 to avoid domination.
• Re-assign public transport demand with new
  Generalised Cost.

18
Multi-class Assignment Output.
19
Express Routes.
20
Before Focusing.
21
First Focusing Step
22
Second Focusing Step
23
Third Focusing Step
24
After Four Focusing Steps.
Dedicated ROW ?
Bus
Mini-Bus
25
Interpreting the results 1

• Blue lines are relatively low volume less that
  100 people per hour in peak Use mini-bus class
  vehicles.
• Green lines indicate sufficient volume to justify
  use of larger buses, in this case, mainly
  standard buses.
• Red lines indicate that consideration should be
  given to dedicated right of way.

26
Interpreting the results 2

• The data gives an indication of where the various
  modes should be employed and estimates of the
  demand levels.
• This gives guidance to
• Rationalisation plans
• Operating Licence Strategies
• New multi-class assignment using final
  generalised costs will give system operating
  conditions for all users integrated transport
  plan!

27
Outcomes Further Development

• Provides a structured approach to public
  transport planning in SA conditions.
• Readily observable results.
• Relatively simple to implement.
• Need to introduce variable demand modelling.
• THANK YOU