Table of Contents
Go Back
Impact of Freight Transport Costs and Pricing
on Logistical Systems
- A System Dynamics Modelling Approach (the SANDOMA Model) —
Valerie GACOGNE
NESTEAR
89-93 Avenue Paul Vaillant-Couturier
F-94250 Gentilly
France
valerie.gacogne@nestear.net
Abstract
The aim of this research is to evaluate the possible impacts of an increase in freight
rates on traffic in a logistical perspective. At the present time road pricing appears as a
key issue for current public policies in Europe in a context of an increasing concern for
sustainable development. However transport costs take part in several kinds of trade-off
decisions that affect a complex logistical system, involving the whole production and
distribution organisation. A complete analysis of their role in the logistical organisation
needs a systemic approach in order to be able to evaluate the impacts of a significant
freight rates increase on traffic. In addition costs are not considered as the only
decisive factor in the organisation, and some qualitative attributes are taken into
account. The SANDOMA model is thus designed to evaluate the impacts of public
policies on freight flows and for a pedagogical use.
Keywords: freight transport, transport costs, public policy, logistics, system dynamics
modelling
Introduction
The work and results presented herein are part of a PhD research carried out at the
French National Institute for Transport and Safety Research, completed in June 2003.
The title of this PhD research is “Impact of Freight Transport Costs on Logistical
Systems with a System Dynamics Modelling Approach — The SANDOMA model”.
Road freight transport costs and pricing appears to be a key issue in current European
transport policy debates. It takes place in a context where congestion, safety, pollutant
emissions seem to be getting worse, and with an increasing concern for sustainable
development, public health and a better quality of life. According to the last White
Paper of the European Commission — “So unless major new measures are taken by 2010
in the European Union so that the Fifteen can use the advantages of each mode of
transport more rationally, heavy goods vehicle traffic alone will increase by nearly 50%
over its 1998 level” (p.13). The enlargement of the Union seems to reinforce this
problem of road freight traffic growth. Firstly acceding States have seen a sharp
increase in their road traffic at the expense of other modes since the beginning of the
90’s, secondly their trades have literally exploded and they are expected to keep on
growing in the next decades.
The following figures remind us of much the land freight traffic has grown in Western
Europe since 1970. Furthermore this growth is only due to road freight transport
(+325%), since railway traffic and inland waterways have remained more or less stable.
As a result their modal share has drastically diminished; indeed rail freight transport
represented a 32% modal share in 1970, but only 15% in 2000. On the contrary, the
road modal share has increased by 23% (reaching a 78% market share).
Table 1: Freight Transport in Western Europe (thousand million of tonnes-km)
1970 1980 1990 2000
Rail 255 242 234 275
Road 439 674 970 1426
Inland 104 106 107 126
waterways
Source: ECMT 2002
All types of measures that are able to encourage a modal shift from road to alternative
modes and also to reduce road freight traffic increase are to be considered. The White
Paper thus offers a set of measures combining road pricing, revitalisation of alternative
modes and investments in the trans-European transport network. The increase of road
freight rates thanks to adequate policy measures is considered as one of the tools to be
placed at the heart of current public policies.
However the role played by transport costs in the whole organisation of freight flows is
not simple. It can be analysed from very different perspectives as in the complex
problem of the modal choice, but it is rarely analysed as a key element in the shippers’
organisation. And yet the logistical organisation of the shippers, which is understood
herein as the whole production and distribution organisation, can directly be put into
question according to some current traffic trends. In other words it seems that the
current freight traffic increase can be partly explained by recent logistical trends. This
research focuses on this issue and how road pricing can be able to reduce freight traffic
growth in a logistical perspective. It appears from the results of this PhD research that
the role played by transport costs is quite complex, as they are implied in many
feedback loops, non-linear and delayed relationships between numerous variables
involved in the logistical system.
The scientific issues in which this PhD research fits and the objectives of the work is
explained with further details hereafter by looking at each term composing the thesis
title. The model that was built using a system dynamics modelling approach is also
briefly presented, and to finish some of the main results coming from the analyses of the
simulations are detailed.
1. Logistical Systems and Freight Traffic Growth
As indicated in the title, logistical systems were studied. First of all, the following two
definitions of logistics are suggested:
- Logistics is currently defined as the ability to make available the right product in
the right quantities at the right place in due time, and at the cheapest cost;
- Logistics aims at optimizing the circulation of physical flows as well as
information and financial flows, from the production of raw materials to the delivery of
the finished product to the final consumer.
This second definition fits in a more recent and modern conception of logistics, in other
words the logistical organisation is conceived as a complex system made up of flow
circulation. With this definition logistics is also understood in its widest sense since it
takes into account the whole production and the distribution organisation; in comparison
to its beginnings, i.e. in the 60’s, logistics was mostly limited to physical distribution.
This is the second definition that was retained for this research. Without going into
details here, we will explain why focusing on logistical systems is important to
understand some of the current traffic trends, and then to evaluate the possible impacts
of a significant increase in freight rates on these trends.
According to recent studies which have been carried out since the beginning of the
90°s', logistical organisations seem to consume more and more transports resulting in
traffic increases, which are usually measured in tonnes-km and tonnes loaded.
At the present time, the way European shippers tend to organise their production sites as
well as their distribution networks would generate more traffics. How?
- By relocating their production units (e.g. in Eastern Europe, North Africa or
Asia);
- By concentrating their production sites;
- By specialising their production sites (for example national multi-product sites
can be replaced by pan-European sites dedicated to one product but with larger
merchandise area);
- By developing the sub-contracting of intermediate stages of the production
process;
- By centralising their distribution structures (with less distribution levels in the
network);
- With the concentration and growth of large retailers;
- By implementing pan-European distribution structures (eventually with the so-
called European Distribution Centres, i.e. serving more than one country);
- By generalising just-in-time practices in the production as well as in the
distribution system.
" See for example BERNADET (1998); MCKINNON & WOODBURN (1993 and 1996).
From all these changes, different kinds of impacts can be expected on freight flows:
- increased international freight transport;
- increased transport links in the production process;
- longer average distances of transport for primary distribution;
- longer average distances of transport in the distribution network;
- but decreased number of distribution links;
- and possible consolidation of freight flows;
- smaller and more frequent shipments;
- more requirements for a high transport quality of service.
To summarize, these changes would increase the average travel distances, and thus
contribute to raising tonnes-km, and also in some cases the number of trips and thus
tonnes-lifted. Furthermore they would favour road transport that is especially more able
to meet quality requirements such as flexibility, reliability etc., as these attributes
weight more and more in these complex organisations.
The two following figures illustrate some of these changes that the logistical
organisation is currently facing. These examples concern the distribution networks.
Figure 1: Distribution structures in Western Europe
From Decentralised To Centralised
Structures Structures
eee ELE
Producers
Production Sites
Distribution Centres
First Level
Distribution Centres
Second Level
Distribution Centres
Third Level
O Production Sites Consumers
A Distribution Centres
Adapted from Bouteiller and Kobler (1998) and Janssen (1993)
This figure indicates that distribution structures tend to be more and more centralised
with less distribution levels and less stocks. The traditional structure which counts four
levels appears to be quite obsolete today in Europe. The one with two levels seems to be
at the present time the most used, with one level corresponding to the national one and
another one to the regional one (or sometimes the first level corresponds to either a
national or a regional one whereas the second one is a local distribution level)’. As to
the most centralised distribution structure, it would thus correspond to direct deliveries
from the production sites to the customers.
The term distribution centre was intentionally chosen as a distribution centre can be
either a warehouse or a depot or even a platform with hardly any stocks that is to say
dedicated to transhipments. A platform can usually replace a storage centre located at
the second or at the third level.
However we must be aware that this decreasing number of distribution centres is also
linked to a move towards supranational structures instead of national ones, with centres
serving more than one European country. This evolution illustrated by the following
? See for example FLEISCHMANN et al. (1998); CRANFIELD CENTRE FOR LOGISTICS &
TRANSPORTATION (1994).
diagram can be compared to a change in the geographical scale that affects the logistical
organisations. It is said that time is a key issue in logistical organisations (with the
circulation of material as well as non material flows going faster and faster), but
spatially organisations have also to be adapted to supranational regions, involving thus
quite complex changes even in the freight flow structure.
Figure 2: Traditional and pan-European distribution structures
Suppliers
Production Sites
Warchouses (finished products)
Main
Distribution Centres
Supranational
Distribution rN
Centres
National E
Distribution Centres :
Regional i
Distribution
Centres fy:
Local
Distribution Centres
‘4 y
Customers
(possible stocks)
oO Production Sites —-> Freight Flows
A Distribution Centres > Information Flows
Adapted from Bouteiller and Kobler (1998)
These changes concern mainly large manufacturers and large retailers, but they are the
sign that European production and distribution systems are currently facing a vast
transformation. Obviously it is linked to the progressive opening up of larger markets
since the end of World War II beginning with the GATT agreements in 1947. This has
lead to intensive trades between countries in Western Europe, and more recently since
the 80’s to the well-known phenomenon of globalisation. These points cannot be
developed in this paper, but it is obvious that globalisation urges manufacturers to adapt
their production system and their distribution networks to larger merchandise areas.
Furthermore it makes emerge areas that correspond to pan-European regions instead of
national ones. Even if the traffic is still closely related to the economic growth, it seems
that nowadays it is necessary to analyse the recent trends in the logistical systems to
understand the freight traffic demand well. This point is the crux of this PhD research.
2. The Role of Transport Costs in Logistical Systems
It appears that road transport costs and rates are directly in question in the recent
logistical evolutions, and this is why an analysis of the impacts of transport costs and
road pricing on the logistical systems appears to be crucial. Transport costs are
questioned in different ways. First of all, road freight rates are incriminated since their
level seems to be quite low regarding other transport modes and the negative external
costs that they are assumed to generate, but are not taken into account. Furthermore as
the road is the dominant mode, road rates are used as a reference according to the White
Paper (p.25), and the current logistical organisations would thus be based on these low
prices. In other words road transport rates would be partially responsible for these recent
trends in logistical organisations which tend to generate more traffic.
It is obvious that transport costs play an important role in logistical organisations, and at
different levels of the system, as they clearly take part in several kinds of trade-offs that
involve production and distribution costs. These trade-offs are related to short-term
decisions that are mostly of tactical concern, but also include medium and long-term
decisions that are thus more strategic. For instance, some operational decisions, i.e.
short-term decisions, will determine the average stock level in a warehouse with the size
and frequency of shipments. An optimum compromise has to be found between
transport and inventory carrying costs, indeed the higher the shipment size is the lower
the transport cost is but the higher the stock level is and consequently its cost. This is
illustrated with the following curve.
Figure 3: Inventory carrying cost versus transport cost according to the shipment
size
Unit
Cost Inventory
4 Carrying
Cost
Transport
Cost
(shipment size) Tonnes
An example of a strategic decision involving transport can be the number and the
geographical area covered by warehouses, so choices that concern the structure of
distribution networks. Indeed transport organisation and costs are directly affected by
these decisions, as the structure of the network will determine the number of the
transport links, the average distances to be travelled, possible flow consolidations, etc.
The logistical organisation implies decisions in which transport costs can intervene, and
leads to determining:
the number and location of production sites;
the degree of specialisation of production sites;
the degree of centralisation of the distribution networks, with the number of
levels in the structure;
the number, the geographical location and the role assigned to the distribution
centres (platforms only for transhipments, warehouses or depots with stocks);
size and frequency of shipments.
As a result transport costs are appreciated differently according to the actors:
from the operators’ viewpoint, transport costs are mostly operating costs which
can be broken down into driver wages, fuel price, truck taxes, etc.
for public policies they are considered as a key tool to reduce traffic increase and
as a possible leverage for modal shift;
from the shippers’ point of view, the role of transport costs appears far more
complex. On the one hand, they correspond to freight rates but on the other hand
they also result from the logistical organisation from which shippers’ decisions
may have a strong influence, as it was explained above. This is why they can be
seen as the origin as well as the result of their logistical organisation. It appears
thus that a complete analysis of their role in the production and distribution
organisations needs a systemic approach, as they are involved in many feedback
loops. They justified the necessity of using a system dynamics approach for the
construction of the model.
3. Presentation of the SANDOMA model
The objectives of the model called SANDOMA? are:
- to highlight the complex link existing between the evolution of freight flows and
current changes in the logistical systems;
- to help understand the role played by transport costs in logistical organisations
as they are involved in different kinds of trade-offs (and as they are considered
as a key tool for public policies).
These two points help evaluate the possible impacts of a significant increase in road
freight rates on the logistical systems and consequently on freight flows. Such an
increase can be induced by adequate transport policies. In the model freight flows are
measured in tonnes-kilometres (tonnes multiplied by kilometres) and tonnes-lifted as
usually in statistics. They are also characterised by average haul lengths, number of
transport links in the networks, shipment size and frequency, possible flow
consolidation, or even the total distance travelled by one product from the production
site to the final consumer. On the basis of these characteristics, further analyses of the
modal choice can be carried out, but it was not in the scope of this research. Modal
choice is one aspect of the shippers’ logistical organisation, and can be considered as
one of the final decisions to be taken at the end of the whole organisation process. For
this research it was essential to focus on logistical trends that deeply affect freight flow
structures, in order to highlight some of the factors responsible for the actual freight
traffic growth.
The role of the model is primarily pedagogical and explicative, as it must help
understand mechanisms which are hidden behind the measures of freight traffic. It also
helps analyse the actual role of freight transport costs in logistical organisations, and
thus the possible impacts of a rise in freight rates on traffic that are so often questioned.
The model was thus designed to be used for the evaluation of public transport policies,
but it can also be applied to a real case with a specific economic sector to be analysed.
The basic model architecture can be sketched as in the following diagram.
* SimulAtioNs Dynamiques pour un Outil d’aide 4 l’intelligence des flux de MArchandises
Figure 4: Basic architecture of the SANDOMA model
Firm
Production
Firm
Distribution
Firm
Logistical System
FIRM DEMAND
ENVIRONMENT (Western Europe)
SUPPLY
The model is intended to simulate the evolution of one firm which supplies one type of
finished product (only one is simulated at the least) in Western Europe, and all its
logistical system at a horizon time of 25 years. In other words its production system and
its distribution system, from the production sites to the final customers, are taken into
account. This firm is constantly interacting with its economic environment that takes
place in a supply system also interacting with a demand system.
The model itself was divided into 16 parts that are obviously interlinked, as follows:
- Western Europe demand;
- Supply and sales;
- Production structure in Western Europe;
- Investments and divestments in Western Europe;
- Production of the firm;
- Firm production facilities;
- Firm product price and costs;
- First level of the distribution structure of the firm;
- Centralised distribution structure (small customers);
- Decentralised distribution structure (small customers);
- Distribution network costs (small customers);
- Distribution structure (large customers);
- Intermediate distribution centres (large customers);
- Firm freight flows in Western Europe;
- Firm transport and inventory carrying costs;
- Imports.
All the organisation of the firm can be simulated, its production as well as its
distribution systems, as the number, capacity and costs of the production sites, the
distribution networks with a more or less centralised structure, the average distances of
transport at the different level of the networks, the number of transport links, the
shipment size and frequency, or even the total distance travelled by each product from
the production site to the customer, etc. Obviously transport costs are involved in many
trade-offs influencing several kinds of decisions to be taken at different levels of the
organisation.
The model counts about 400 variables, and it is based on the following feedback loops
which appeared to be quite complex as dynamical hypotheses.
Figure 5: Basic feedback loops of the SANDOMA model
Production Sites
Concentration
. SC)
Average Distances
Policy Pricing
+
Stock Level
.
2)
Distribution Structures
nee Transport Costs
Centralisation P
+ é)
Inventory Carrying
Costs
This diagram is related to freight flows from the production sites to the final customers.
It presents the basic feedback loops in which transport costs are involved. For instance,
a more centralised distribution network tends to increase distances and consequently
transport costs, which in turn influence decisions related to the distribution structures. In
addition a more centralised structure with less distribution centres tends to diminish the
global level of stocks in the network. It appears then that in some cases the average size
of shipments is raised to counterbalance the stock reduction in order to decrease the
shipper’s transport costs (as road freight rates tend to decrease with the shipment size).
Without going into further details, these few examples taken from the previous diagram
show some of the most important feedback loops of the model. They can give an insight
11
into the complexity of the role of transport costs, of which effects on the logistical
organisation are neither direct nor linear.
4. Scenarios and Results
Results can be clustered according to two different issues which were analysed by using
different sorts of scenarios. The first one concerns the impacts of current logistical
trends on freight flows, and more especially the consequences of the concentration of
the economic activities and of specialisation of sites on traffic. The second one deals
with the valuation of the possible impacts of a significant increase in freight rates on
logistical organisations and consequently on freight flows.
Secondly, the model was used in two different manners. It was applied to a specific case
which is the production and the distribution of eggs in Europe. This case was taken
from reality with actual data from the economic sector, and also directly given by
professionals and the firm which was studied in detail. Results of the simulations
validated the model and provided a good understanding of the current organisation; it
also gave insights into the possible evolution of the organisation of the economic sector.
The model was also applied to different types of products with well-defined
characteristics (added value, weight, demand, etc.) quite realistic but not coming from a
real case as for the eggs. These kinds of simulations were mostly designed for a
pedagogical use of the model.
Concerning the effects of the concentration and specialisation of the activities
(production and even distribution centres), the results confirm some analyses that
assume that they tend to increase significantly haul lengths, and thus tonnes-km
measured in statistics. However the model also shows how important the flow
consolidation is, due to the reduction of the transport links inside the distribution
networks (simultaneously with the concentration of activities). Furthermore this
consolidation tends to raise the average shipment size which can thus partially or
completely offsets the transport cost rise resulting from longer distances.
In the model, large and small customers have been distinguished. Actually large
customers correspond to large retailers or wholesalers; they have very different types of
organisations (and costs) in comparison to a small retailer or a small firm for at least
two reasons. Firstly as they are able to put pressure on some of their suppliers, they are
also able to influence the logistical organisation of their suppliers to make them meet
their requirements (fast replenishments, bigger shipments, flexibility, etc.). Secondly as
large manufacturers they have the ability to concentrate and consolidate large freight
flows, in that way they are able to make substantial transport cost reductions. Large
retailers have already made a great decrease in stock levels and inventory carrying costs
thanks to a high stock turn-over. The next step is obviously to keep on reducing
distribution costs by decreasing transport costs thanks to flow consolidation. The
simulations clearly showed that the transport cost rise due to longer distances cannot
compensate for the transport cost reduction induced by flow consolidation, which is
12
allowed for by concentrating activities and the huge volumes generated by large
shippers.
Concerning the possible impacts of an increase in freight rates, which would come from
public policies via taxes or specific social regulations for example, it seems that even if
it is significant (more than 60% over 25 years) it is not really able to influence the
concentration of activities and consequently the trend in haul length increase. This
assessment can be explained in two ways:
- Transport costs are a major factor in the logistical organisation for some
products that are of low added value for example (as in the case of the
production and distribution of eggs that is also a perishable good). As a result
the whole organisation of such product tends not to concentrate a lot the
distribution sites (the break-even point depends on many variables depending on
the characteristics of the product but also on the production system, the volume
of sales, etc.), and to keep the average haul length quite short. Consequently,
whenever transport rates are raised it seems that there is no flexibility in the
system to adapt itself to this change. Depending on the case study, if possible the
frequency of the shipments is reduced and thus the size is increased leading also
to higher stocks. But most of the time the average shipment size is quite high for
these products, and sometimes already corresponds to a full truck load. So
another response is a lower price margin and eventually a price increase
(depending also among other of the economic environment and the competition).
- At the contrary, products for which transport costs do not seem to play an
important role in the logistical organisation are not very sensitive to a sharp rise
in rates. Usually just-in-time practices are widely used and the most frequent
response to such a rate increase is again larger shipments with a lower
frequency. But if it was the shipper’s advantage to concentrate activities, which
implies longer haul length, most of the time it is still his advantage not to change
such an organisation when freight rates are raised. This result was not really
intuitive but appears logical with the simulations of the model.
Conclusion
SANDOMA seems to be the first model in Europe that looks at freight transports using
system dynamics modelling with a micro-economic approach. This choice was made to
highlight mechanisms that underlie freight traffic trends under the influence of the
increasing complexity of the logistical systems. It was also made to understand the
actual role of transport costs in the logistical system, since they are considered as a key
issue for the Common Transport Policy.
Obviously not all the results can be developed in this paper, but some of them have been
underlined. Even if the consequences of public policies aiming at increasing freight
transport costs do not seem to lead to the expected results, we must keep in mind that
13
the concentration of activities also allows for flow consolidation. This trend tends to
raise haul length and thus tonnes-km, but it also has a strong impact on the
consolidation of flows, and yet this point cannot be statistically measured. However
such a trend is of great importance as it changes the structure of the flows, and needs to
be further analysed from different perspectives including environmental, and the use of
vehicles (types, average load, etc.).
We hope to have in the future the opportunity to improve the SANDOMA model and to
use it for further analyses of freight traffic in Europe. Systems dynamics modelling
proved to be a powerful tool to help understand complex logistical trends and to analyse
the role of transport costs that are key issues for current transport policies.
14
References
A.T. Kearney, 1993. Logistics Excellence in Europe. Bruxelles: European Logistics
Association, 1993. 58p.
A.T. Kearney, 1999. Insight to Impact. Bruxelles: European Logistics Association,
1999. 35p.
Bernadet, M., 1998. Les transports en France - Repéres chiffrés tirés des comptes des
transports de la Nation. Paris: Economica, 1998. 143p.
Bouteiller, R., Kobler, R. A., 1998. “Strategic EuroLogistics Design”. In Advances in
Distribution Logistics. Lectures Notes in Economics and Mathematical Systems. Berlin:
Springer, 1998. pp.3-25.
Cranfield Centre for Logistics & Transportation, 1994. Logistics Futures in Europe — A
Delphi Study. Cranfield: Cranfield University, February 1994. 109p.
ECMT, 1994. Internaliser les cotits sociaux des transports. Paris: OCDE, 1994. 21 1p.
ECMT, 1997. Les nouvelles tendances de la logistique en Europe, rapport de la 104°"
Table ronde d'économie des transports, Paris, 3-4 octobre 1996. Paris: OCDE, 1997.
189p.
ECMT, 2002. Evolution des transports 1970-2000. Paris: OCDE, 2002. 67p.
European Commission, 1995. Livre Vert — Vers une tarification équitable et efficace
dans les transports. Bruxelles: décembre 1995. 58p.
European Commission, 2001. White Paper — European transport policy for 2010 : time
to decide. Bruxelles: October 2001. 120p.
Fleischmann, B., et al. (edited by). 1998. Advances in Distribution Logistics. Lectures
Notes in Economics and Mathematical Systems. Berlin: Springer, 1998. 535p.
Janssen, B.J.P., 1993. “Product Channel Logistics and Logistics Platforms”. In Europe
on the Move. Avebury: Aldershot, 1993. 7p.
Lamy, 2002. Guide des cotits de transport — prix de revient et tarification routiére de
référence. Paris: Lamy, avril 2002. 281p.
McCann, P., 1998. The Economics of Industrial Location- A Logistics-Costs Approach.
Berlin: Springer, 1998. 228p.
15
McKinnon, A.C., Woodburn, A., 1993. “A logistical perspective on the growth of lorry
traffic”. Traffic Engireering +Control, October 1996, Vol.34, n°10, pp.466-471.
McKinnon, A.C., Woodburn, A., 1996. “Logistical restructuring and road freight traffic
growth - An empirical assessment”. Transportation, 1996, n° 23, pp.141-161.
OCDE, 1996. Logistique intégrée avancée pour le transport de marchandises. Paris:
OCDE, 1996. 199p.
16
Back to the Top
