Nicolas Lebelle

1
Energy consumed in freight transport first
results from the shipper and operator survey

• Nicolas Lebelle
• Philippe Marchal
• Christophe Rizet

2
Structure of this presentation

• Part One an overview of SOS
• Objectives, structure and sample of the survey
• Geocoding aspects
• Illustration of general results generation
• Part Two energy consumption
• Computing energy consumption
• Energy efficiency of road freight transport

3
Part one an overview of INRETS Shipper
Operator Survey (SOS) - tracing the shipments
in Europe
4
1 Objectives of the survey

• We have good statistical data for each single
  mode, derived from the transport sector but
• no relations between the modes (no information on
  multimodal transport) and
• no link with economic activity
• SOS makes this information, by tracking the
  shipment along the chain
• A first Shipper surveys in France in 1988 and two
  limited surveys in 1999 (NPDC, Mystic)
• 2004 SOS includes several improvements and a new
  objective quantifying energy consumption

5
The 4 levels of the survey
6
Tracing the shipment along the chain
7
Geographical aspects pre-geocoding

• Before the field survey
• A list of pre-geocoded places in the CAPI limits
  the localization errors and the burden a draft
  list based on NIMA database (worldwide coverage
  of the survey)
• Problem presence of multiple values in the
  names
• (Ex 3 Frankfurt in Germany)
• An automated process was developed using another
  database of main cities with population data
• Ex
• Frankfurt /45km/ Nurnberg
• Frankfurt /82km/ Berlin
• Frankfurt /0km/ Frankfurt

8
Geographical aspects additional geocoding

• After the field survey
• Why an additional geocoding ?
• missing or erroneus coordinates
• mis-spellt names, inconsistency with the initial
  CAPI database
• How ?
• Spell checking functions similarity tests
  between names

9
Geographical aspects transport chains
validation and distances estimation

• a) Detection of "suspect" shipments, by checking
  
• Consistency between shipments and legs (final
  destination for shipment destination for last
  leg)
• Legs chaining (destination for step i origin
  for step i1)
• b) Additional checking based on distances
• Shipments distance classes and countries
• Legs distance classes and modes
• c) Road distances estimation
• A standalone tool for visualization and
  processing

10
Geographical aspects computational tool
11
OPTIMIZING THE SAMPLE

• 2 Objectives
• Sufficient number of shipments for non-road
  modes ( for the north region)
• Improve the accuracy of the results
• Sampling method the 2 steps
• Sampling the firms from an exhaustive list
  oversample the firms using non-road modes and in
  the North (Activity localization) Strates
  based on the number of employees to improve the
  accuracy
• Sampling the shipments in the CAPI, 3
  shipments are randomly chosen among the last 20
  shipments The probability of being selected is
  weighted in order to adapt the sample to our
  objectives.

12
Computing energy consumption

• 1) At the leg level
• Energy per vehicle leg
• Evl (distance empty) f(vehicle type, total
  weight)
• Energy per shipment leg
• Esl Evl (shipment weight / total vehicle
  load)
• 2) At the shipment level
• Energy per shipment Es SumEsl
• 3) At the company level
• Energy per Cy Ec Sum Es

13
Part 2 Overview of first results

• The sample
• 2 962 establ. over 10 (or 6) employees
• 10 462 shipments traced of which 9 742 complete
  transport chains
• 27 069 operators (intervenants) ??
• 20 074 legs

• The population
• 69 256 establ shippers
• 738 millions of shipments

14
Traffic generation / activity Yearly tonage
(1000 t.) / estab.
15
Traffic generation / size of establ. Yearly
tonage (1000 t.) / establ.
16
Traffic generation / employee Yearly tones /
employee
17
Energy consumption first results
18
Computing energy consumption

• 1) At the leg level
• Energy per vehicle leg
• Evl (distance empty) f(vehicle type, total
  weight)
• Energy per shipment leg
• Esl Evl (shipment weight / total vehicle
  load)
• 2) At the shipment level
• Energy per shipment Es SumEsl
• 3) At the company level
• Energy per Cy Ec Sum Es

19
Average energy efficiency for road goe/tkm
20
Energy efficiency per shipment goe/tkm is
very scattered (2002 results)
21
Next steps on SOS and energy

• A very powerful tool for research, linking
  information on the shipper, the shipments, the
  operators, the transport logistic services
• Modeling the influence of logistical choices on
  energy consumption and energy efficiency
• 4 steps model energy consumption
• Clusters of establ. in which a direct estimate of
  energy would be possible