Freight transport modelling - an approach to understand demand and use of transport energy

1
Freight transport modelling - an approach to
understand demand and use of transport energy

• Annecy, May 26th, 2008
• Ole Kveiborg and Jean-Louis Routhier
• Institute for Transport Laboratoire d'Economie
  des
• Technical University Transports
• of Denmark Lyon, France
• ok_at_transport.dtu.dk jean-louis.routhier_at_let.ish-
  lyon.cnrs.fr

2
Freight transport modeltypology
Energy consumption
3
Freight transport modeltypology

• Production and consumption. Deals with decisions
  about the location of the production and
  consumption, type of products and the volume.
• Spatial distribution of trade. The relation
  between production and consumption locations has
  to be established. To what region will the
  produced goods be sold (sales) or from what
  region are the goods bought (sourcing)?
• Logistics on firm level. Decisions about the use
  and the location of inventories and about supply
  chain management.
• Mode and route choice. What transport modes and
  what types of vehicles will be used for
  transporting the goods on which overall route
  (not necessarily linked to a detailed network.
• Transport logistics. The utilisation of the
  vehicles, load factors, empty running and similar
  decisions made by the transport provider.
• Networks and assignment. The actual allocation of
  vehicles onto the physical road network.
• Energy and environmental consequences.

4
Aggregate and disaggregate approaches

• What is the relevant policy question?
• Future flows in spcific corridors
• National energy consumption
• Determining macro drivers for future flows
• Choose appropriate model
• Detailing on the focus area
• Not one model can answer all questions

5
An aggregate approach

• The influence of different macro drivers
• Kveiborg et al. (Arcueil, 2005 Torino, 2007)
• A decomposition of freight traffic (and
  transport) growth on
• Economic activity
• Physical content of production
• Handling (logistics in an macro sense)
• Load and length
• Empty runs
• Decoupling (or coupling) is a consequences of
  adverse influencing major drivers

6
An illustration of modelling approach
Urban freight modelling and energy consumption
7
Urban goods modelling and energy issues
Several submodels...
Demand modellingDelivery and pick-up generation
for the total industry in the urban area
Vehicle flows modellingVehicle movement
generation (veh.km,
Vehicle energy consumption modellingtype of
vehicle, motorization, speed, loading,
acceleration,
...To explain the goods vehicle energy
consumption in the urban area...
Calculation of the impact of the urban logistics
on vehicle flows
types of vehicles (LGV, HGV), engine
specifications
logistical choice (routing, packing, tracking)
location of firms, warehouses and consumers
consumer behaviours (home deliveries,
e-commerce,)
economic valuation
...Policy Oriented (PO) Models to help
forecasting and decision making.
8
Methodological commitment of urban freight
modelling
Inside the Town the organisation in tours is
dominant
- The O/D matrix of goods is different from
the O/D matrix of vehicles
- Vehicles have very different sizes and freight
volume
- Packaging are very different
One quantity of goods may be delivered by
different types of vehicles, of way of
organisation etc.
For a given commodity, the flows of vehicles are
determined by factors exogenous to transport.
The commodity gravity model is failing
The movement an efficient unit of observation
a good knowledge of the generators
a good description of the deliveries and pick-ups
9
Urban goods modelling and energy issues
Few PO UGM models to answer the issues
An example The Freturb Model (COST Arcueil and
Berlin 2005)
The total freight in the city Three modules
An effort of data collection
Specific surveys (4500 establishments 2,200
drivers in three different size towns)
Pick-ups and deliveries model(comodity flows
between establishments)
Town management module(raw material and goods
works, urban networks, removals)
Specific surveys
Purchasing trips model (last mile by the
consumer)
Household trip survey and specific surveys
Output Indicators
number of deliveries and pick-ups per economic
sector
number of vehicle.km for different types of
purpose and types of organisation
number of on street parking (congestion)
fuel consumption and CO2 emission (per day, per
inhabitant, per job)
10
Urban goods modelling and energy issues
Output of Freturb

• Street occupancy, Traffic and Energy
  consumption
• per vehicle
• per industry sector
• per traffic segment
• according to the management mode (hauliers, own
  account)
• energy balance between goods movement and
  trips for purchase in a town.

• Examples

11
Urban goods modelling and energy issues

• Good Input Variables for Simulation
• location of firms, trade centres, warehouses
  and consumers
• logistical choices (management mode,
  co-operation,..)
• consumer behaviours
• interaction between commercial transport system
  and individual trips for purchase
• interaction demand(need of goods)-supply(transpo
  rt operating)

12
Urban goods modelling and energy issues

• Brakes on PO-UGM modelling
• it is difficult to consider the total
  transport activity
• lack of data (costly, lack of interest)
• models are not widely policy-oriented
• O/D distribution is difficult to calculate in
  urban areas
• connection with the upper scope models
  (regional, national) difficult to compare
  methods and results of different modelling
  approaches

13
Urban goods modelling and energy issues

• Some recommendations
• Effort in developing and harmonising data
  collection
• Improvement of the integration of different
  scopes of models,
• to harmonise the space and time units
• to list and analyse the main input exogenous
  and endogenous variables
• to improve the efficiency in terms of
  prediction of energy consumption.

14
General recommendations and future unsolved issues

• Links between scales of interest
• Local policy and global policy analysis require
  different types of models, but links and
  consistency between them is often lacking
• Local O/D matrices difficult to obtain and how
  are they related to regional/national O/D
• Data on commodity based production and freight
  transport (vehicle approach) difficult to combine
• Knowledge of logistics and inclusion in models
  still in its infancy
• Data focussing on linking demand with e.g. supply
  chains and final use