A

TERM PAPER

ON

SUPPLY CHAIN MANAGEMENT IN SUGAR INDUSTRY

SUBMITTED TO

Dr. P R SHARMA

PROFESSOR

GSIB
BY

G. Arun Kumar

1226210109
MBA (IBF)

SECTION-C

GITAM UNIVERSITY

GSIB

VISAKHAPATNAM
 SUPPLY CHAIN MANAGEMENT IN
SUGAR INDUSTRY
Abstract:
Greater competitiveness and deregulation in agribusiness and the food industry require new
forms of co-ordination between farmers and their clients to increase the efficiency and
profitability of the supply chain. Various co-ordination processes are used by processing firms to
control the quantity and quality of their raw material. In the sugar industry, millers plan their
cane supply to ensure that the mill operates at optimum capacity throughout the entire season.
They may also take into account variations in cane quality within the supply area and at different
times during the season to maximize sugar production.

These decisions will impact on the choices growers make with regard to their harvest capacities
and management and, depending on the cane payment system in place, on their incomes as well.
Other stakeholders in the supply chain, such as contractors and hauliers, also directly affect its
management and results. Any modification to the structure of the supply chain or the
management rules should take into account stakeholder strategies and ways of operating.

A modeling approach based on two complementary models has been developed to simulate on a
weekly basis the planning and operation of mill supply throughout the season. The first model
compares weekly and total sugar production for a season. The second model focuses on the
simulation of logistic chains, and enables the impacts of technological and structural changes on
daily harvest and transport capacities to be assessed. Both models can be used to support
discussion and negotiation between growers and millers regarding evolutions in the supply chain
management.

Introduction:

Escalating market instability calls on firms to be increasingly reactive and flexible, and implies
the development of new forms of industrial organisation in agro-food chains. In this context, the
control of products from supply areas to processing plants, and from plants to markets, is a
critical factor of efficiency. Because numerous elements interact within the supply chain, such as
industrial capacities, production risks and diversity of farm structures, finding organisational
solutions that satisfy the objectives and constraints of all stakeholders is not a simple process.

The sugar industry faces such co-ordination problems, especially when a large number of cane
suppliers are involved. Millers organise their cane supply to ensure regular mill operation
throughout the entire season in accordance with milling capacity, and may also take into account
variability in cane quality, to maximise sugar production. Decisions made by the millers impact
on the choices growers make regarding their harvest capacities and management. Depending on
the cane payment system in place, these decisions could affect grower incomes as well.Other
stakeholders, such as contractors and hauliers, also directly affect supply chain management and
results.

This paper investigates modelling approaches and support tools that could be valuable in
negotiations between firms seeking organisational solutions to problems. The first part of the
paper briefly reviews theoretical frameworks used in chain analysis, and suggests a modelling
approach to flow of material between firms. The approach provides negotiation support to
stakeholders involved in the supply chain by highlighting the impact of individual strategies on
the efficiency of the chain as a whole. The second part describes how this approach has been
implemented in the sugar industry. Practical applications in La Réunion and South Africa are
presented at this Congress in two complementary papers

Review of Literature:

Some research work has shown that joint decision making between several firms yields higher
profits for the whole chain in the case of a distribution channel. However, such problems are
difficult to solve in practise. Different theoretical frameworks have been proposed to deal with
this type of problem, either in the field of economic analysis of vertical relationships, or in the
field of Management Science and Operational Research.

Co-ordination within the supply chain

Co-ordination between firms within the supply chain involves four processes:

▪ definition of contracts and limitations

▪ management of supply process

▪ information management

Contract and limitation issues have been studied mainly in the field of economic theory. For
example, incentive theory aims at designing optimal contracts to induce co-operation in a world
of asymmetric information plagued by the opportunism of the agents .Applications of this theory
look for pricing rules and incentives. Limitation analysis focuses on the impact of vertical
restraints on chain efficiency, and their compatibility with antitrust policy.

A large number of studies have been based on these two theoretical frameworks. Some describe
modes of co-ordination between firms, such as strategic alliances between agro-business firms
(Sporleder, 1990). Others highlight the role of information technology and information
asymmetry in encouraging vertical integration within the supply chain.

Some of this work has led to a modelling approach aimed at characterising the level of inter-firm
integration within a chain. For instance, Franck and Henderson (1992) focus on how transaction
costs impact on food industry vertical linkage. They show that these costs are critical factors for
firms that are vertically co-ordinated by non-market arrangements. Focusing on vertical
relationships in the Champagne wine chain, Giraud-Héraud et al. (1999) analysed the market
between grape producers and wine merchants. Their model simulates the impact that the
economical behaviour of producers and merchants has on the efficiency of the chain.

Optimization approach:

In the field of Operational Research and Management Science, there have been many
investigations into the value of co-ordinating two functions within a firm (to build a production
and distribution plan, for example) or several firms within a chain.

The main purpose of these studies is to find efficient production schedules, product delivery and
inventory pricing in the face of unstable demand. Usually, modelling is used to find the plan that
provides the best co-ordination between marketing and manufacturing.
For instance, Chandra and Fisher (1994) study the co-ordination problem by comparing two
approaches: (i) where production scheduling and vehicle routing problems are solved separately,
and (ii) where they are co-ordinated in a single model. They combine the number of products and
retail outlets, the cost of set-ups, inventory holdings and vehicle travel. de Groote (1994)
addresses the problem of flexibility of manufacturing processes in relation to marketing. He
analyses both product variety and process flexibility, concentrating on changeover costs and
inventory costs. This leads to insights into manufacturing systems and .just-in-time. production.
Some studies related to this field of research have been carried out in the Australian sugar
industry.
An optimisation model was developed at the mill area level to maximise the sugar yield and net
revenue of the chain by defining suitable harvest dates and crop cycle lengths for a range of
production units (Higgins, 1999; Higgins et al., 1998). This model includes cultivation, harvest,
transport and transformation costs, and provides an optimum solution for harvest scheduling,
given the structure and capacity of the supply chain. The model appears well suited to the
Australian situation, since the mill and its suppliers are closely integrated. Indeed, the growers
still harvest their own cane - but it is the mill that schedules the arrival of the containers used for
cane transport. As harvest is almost totally mechanised, and cane deterioration thus all the more
rapid, the growers have the incentive to follow the schedules imposed by the mill.
Inter-firm modelling as a support for decision making
Theoretical economic frameworks are useful for analysing and understanding the strategic
interactions between stakeholders (information asymmetry, market powers), but they miss the
technical aspects of the observed phenomena (inventory costs, delay costs, production systems).
These approaches focus more on the way the total chain value is shared between stakeholders
than on ways of increasing that value (Soler and Tanguy, 1998).
The reverse situation occurs in the field of Operational Research and Management Science,
where technical issues are highlighted, and interaction between stakeholders receives least
attention. Management tools are often based on an implicit organisation, which is far from being
realistic; i.e. there is no communication problem and no conflict of interests, and uncertainty is
taken into account in a .probabilistic. way.
The alternative approach described in this paper aims to go beyond these limitations by
combining the potential of both theoretical fields.
This approach, which has been applied to the wine industry as well as the sugar industry (Gauche
et al., 1998), is based on the following principles:
1. It will be more useful for stakeholders to consider ways of increasing the total value of the
chain, rather than engage in altercation over the distribution of the existing value. The first
issue is thus to investigate the possibility of developing co-operative strategies. Value sharing
will be the second stage in the process.
2. This position leads to designing a collective plan for the whole chain, and requires stakeholder
commitment and a description of individual expectations. The link between whole-chain
strategy and the way in which flow of goods is managed between stakeholders is therefore of
particular significant.
Mechanisms regulating quantities and prices will be designed when each stakeholder has
accepted both the whole-chain strategy and its impact on flow management.
These agreements pave the way for controlling the common strategy. Critical factors
governing the strategy.s success are negotiated between firms, either within an inter-
professional body such as the South African Sugar Association, or through a classic market
relationship between firms.
This approach confers a major role to the development of planning simulation models, rather
than sophisticated forecasting tools, optimal scheduling systems or incentive devices. The
models are designed to assist stakeholders in reorganising the planning and operation processes,
and feature a number of key characteristics consistent with their use in planning when it is
considered as an interactive process between stakeholders (Ponssard and Tanguy, 1993). This
approach assumes that a certain degree of uncertainty is a priori collectively admitted. It also
assumes limited communicative capacity and a certain amount of inflexibility in the transfer of
resources from one stakeholder to another.
The support process does not aim at formulating new policies by using a model which the
stakeholders regard as a .black-box., but aims instead at designing new benchmarks, first by
considering any and all conflicting points of view about future plans and then by investigating
mutual readjustment of individual proposed positions.
Through this approach, modelling provides support tools that aim at facilitating negotiations
between stakeholders (Tanguy, 1989; Nakhla and Soler, 1996):
1. Describe and discuss the implicit rules governing technical stakeholder interfaces. Lack of
clarity in these rules, and a consequent misinterpretation of their content, can lead to
breakdowns in co-ordination and have serious consequences.
2. Stimulate collective investigation to find more efficient rules, and to share the overall
economic stakes associated with the governance derived from these rules.
3. Establish monitoring and control of the new arrangements determined by the rules.
The next section of this paper shows how this theoretical background has been put into practice
within the sugar industry. The conceptual model described here is based on a research project
conducted since 1996 in La Réunion Island. Its design has been confirmed by a similar study
carried out in South Africa since 2002.

Application of the model to the sugar industry

In the cane or beet sugar industry, co-ordination between growers and millers focuses on the
organisation and process of the mill supply. Decisions made by millers regarding mill capacity,
the location of mill and transloading centres, and delivery allocations, will impact on the choices
made by growers regarding mechanisation and harvest management. In turn, decisions made by
growers regarding variety selection, harvest capacity and work organisation, will impact on
milling efficiency. Poor cane quality will reduce crushing capacity, while irregular deliveries will
disrupt the continuity of mill supply. Intermediate operators involved in cane flow management,
such as harvest contractors and hauliers, will also affect the supply process. Total sugar
production at mill area level thus depends on the efficient functioning of these technical
interfaces, as well as on each stakeholder’s management processes.

To increase the total value of the chain its various components, from the cane fields to the mill,
must be taken into account.
Issues that should be considered are:
1. How to match capacities (industrial and agricultural) with an unstable tonnage of cane at the
industry or mill area level: number of mills, individual capacities, location of mills in existing
or potential supply areas?
2. How to organise the mill supply area in order to transport cane flows from the fields to the
plant: location and capacity of transloading centres, role of hauliers and contractors?
3. Which planning and operation rules would be efficient and in line with the objectives and
constraints of each stakeholder: season length and dates, delivery allocation and flow
monitoring
A modelling framework consisting of two components was developed to address these issues,
with interaction between the miller and the growers that supply the cane. The model is based on
the above theoretical background and on an analysis of real mill supply organisation in La
Réunion and South Africa.
These two cases are similar, in that
(i) a large number of growers supply each mill, and
(ii) similar tasks are carried out as the cane moves from the growers to the mill. The difference
between the two cases is in the way these tasks are managed by the different stakeholders
.
The modelling framework provides a comprehensive picture of mill supply chain management,
with particular focus on interaction between stakeholders along the chain, and simulates and
compares supply scenarios by assessing the impact they have on each stakeholder and on the
chain as a whole.

The first modelling component looks at supply management over the entire mill area, and is
called the strategic model as it addresses mid-term strategic issues such as the relocation of mills
and transloading centres, investments in industrial and agricultural capacities and new rules
governing delivery allocation. It models the planning and operation of a crushing season on a
weekly basis, based on a simplified representation of both cane flows and the stakeholders
involved in the process.

The grower level is fairly simple, as it is not possible to take into account the harvest and
transport constraints of each individual supplier. Harvest, loading and transport capacities are
thus combined into an aggregated variable, which gives an estimate of potential capacity that can
be used by a group of growers or a cane production area to convey a given amount of cane
during a week.
The second modelling component, the logistic model, details the way cane is conveyed from the
fields to the mill on a daily basis. It aims at assessing the impact that logistic constraints such as
harvest, loading and transport capacities, have on mill supply management. As seen in Figure 2,
the two components are linked conceptually. The feasibility of aggregated capacities used in the
strategic model can be addressed in the logistic model, while the capacities of new logistic chains
can be simulated in the logistic component before being implemented in the strategic model.
Strategic model

The strategic model is made up of two sub-modules, the first for supply planning and operation,
and the second for cane processing (Figure 3). It is based on a three-level representation of the
supply area, including mill, intermediate operators such as hauliers and transloading centres, and
production units (PUs), and simulates a crushing season on a weekly step basis, according to the
cane flow paths between these three levels, their given characteristics and certain planning and
operating rules.

Unplanned events such as mill breakdowns or delivery shortfalls, and joint rules of adjustment,
may be introduced into a scenario. For example, the non-delivered tonnage allocated to a given
PU can be transferred to other PUs linked to the same intermediate operator, as long as this does
not exceed the operator’s handling capacity for the week. If it does, the extra tonnage will be
transferred to another intermediate operator, who will re-allocate it to his PUs, or it will be
carried over to the next season. Such events and adjustments will impact on final sugar
production, as sucrose curves vary from one PU to another.

The mill is described in terms of its weekly crushing capacity and its sugar extraction efficiency.
Weekly crushing capacity is calculated from elements that take into account mill operation and
work organisation, such as hourly crushing capacity, weekly working hours, maintenance time
and frequency of breakdowns. Both hourly crushing capacity and sugar extraction efficiency
depend on the quality of the cane during a given week. For example, South African millers
calculate a quality-dependent crush rate by taking into account brix, fibre or non-sucrose
handling capacity, sugar lost during processing and cane quality figures forecast from historical
data.
Production units are defined as individual elements considered as homogeneous from a capacity
and cane quality point of view, as well as the planning and operating rules that the mill applies to
them.
Production units are thus described in terms of (i) cane quantity (cane area and yield), (ii) a
quality curve and (iii) an optimum flow rate that includes limitations for harvest and transport to
a given transloading zone or to the mill.

PUs are a group of growers located in a given zone within the mill supply area. Location governs
the distances between the production areas and the mill, which impact on transport capacity, and
climatic variables such as rainfall and temperature, which impact on cane yield and sucrose
curves. Harvest, loading and transport capacities are assessed either by conducting farm surveys
or by processing data collected for each consignment at the miller’s weighbridge. These values
are then aggregated to evaluate the potential capacity that will be used in the simulation.

Each PU is assigned to an intermediate operator, who collects cane from different units and
transfers it to another operator, or takes it directly to the mill. These operators are described in
terms of maximum weekly transfer capacity that includes cane handling capacity, storage
capacity and limitation for final transport to the mill.

Based on the definition of the supply organisation, the planning and operation module aims at (i)
determining a planned delivery schedule and (ii) simulating this schedule, with unforeseen
events being introduced into the scenario.

The weekly delivery schedule is defined according to the following procedure:

1. Cane tonnage to be delivered by PUs is estimated from the cane area and the yield.

2. The length of the milling season is calculated by dividing total cane production by mill
crushing capacity.

3. The closing date of the season is defined and the starting date is calculated according to the
season duration.

4. Weekly delivery rights are allocated to PUs, according to particular rules, e.g. rateable
deliveries over the entire season or variable deliveries during specific harvest windows.

Simulations run by this module provide a weekly delivery schedule for the entire mill supply
area, as well as for each PU. The cane deliveries are then transformed into sugar production
through the processing module, taking into account PU sucrose curves and milling efficiency.
The amount of sugar produced weekly and annually is calculated and used as an indicator when
evaluating and comparing scenarios.

Other indicators can be used, such as (i) the difference between potential and actual capacity
along the chain, which highlight under- or over-capacities, (ii) the amount of cane carried over to
the next season and (iii) mill supply irregularities, which show adjustments that should be made
to cater for unforeseen events.

This model provides a way to:

1. characterise the links between the mill supply organisation and the planning and operating
rules governing supply at the mill area level

2. evaluate how these links impact on supply chain efficiency

3.support stakeholder discussion about changes in supply management by comparing simulated

The model has been used in La Réunion and in South Africa to address alternative mill supply
management, which takes into account the variability in cane quality within the mill supply area
(see Lejars et al., 2003; Guilleman et al., 2003). A specific computer program, MAGI, developed
from this conceptual framework, facilitates simulations and transfer of the approach to various
potential users .

Logistic model

Modelling of logistic chains has already been carried out in the sugar industry, to investigate the
reduction of harvest-to-crush delays (Barnes et al., 1998, 2000; Semenzato, 1995) and matching
harvest capacity to mill crushing capacity (Arjona et al., 2001). These models describe precisely
the paths followed by cane consignments and the machinery used from fields to mill, with the
aim of pinpointing potential bottlenecks and their impact on the supply chain.

A similar methodology was used in La Réunion to evaluate the impact of various changes in the
supply chain (alternative means of transport, closure of transloading centres.) on cane harvest
and transfer capacities. While the strategic model takes into account the entire supply area, the
logistic model simulates a given logistic chain on a daily basis, leaving the rest of the supply
organisation the same. These principles allow a detailed description of the tasks involved in
conveying the cane from the fields to the mill via a transloading centre. For example, PUs are
characterised by their cutting, loading and transport equipment (size, number of machines,
hourly capacities), their work organisation (number of workers, working schedule), the distance
to their allocated transloading centre and their transport speed, depending on the type of road
they use. At the transloading centre and mill levels, cane flows delivered both by the logistic
chain being studied and by the remaining suppliers are simulated in order to calculate queuing
delays.

An industrial simulation tool has been used to model and simulate cane flows associated with a
given logistic chain. The modelling structure is based on three modules: (i) growers’ cane fields,
(ii) transloading centre and (iii) millyard (Figure 4). The modules are linked by cane transport
from one level to another. For example, a cane trailer is loaded in the field, taken to the
transloading centre to be unloaded and is then sent back empty to the field. Rules of work
organisation such as cutting time or opening-closing time of transloading centres, are also taken
into account.

Using the logistic model in conjunction with the strategic model, issues can be addressed in
detail. The logistic model is used mainly to assess the impact of (i) restructuring the supply
organisation, e.g. by closing or opening transloading centres to modify distance from fields, (ii)
introducing new equipment at PUs, transloading centres and at the mill, or (iii) changing delivery
allocations to match the harvest capacities of growers.

This model was tested in La Réunion to assess the value of grouping growers around bigger
loaders and trailers, to reduce queues at a smaller number of transloading centres. The results
obtained by simulation show that loader utilisation decreases when the distance between fields
and transloading centres increases (Figure 5). This negative impact can be partially corrected by
increasing the number or capacity of trailers, keeping the tonnage delivered the same. Closing
some transloading centres means grouping the more distant growers around fewer, bigger
trailers. Simulation allows general results to be quantified and discussed in more detail.

Conclusion

The modelling framework described in this paper aims at providing stakeholders with
information that will enhance discussion and negotiations with regard to the design of an
efficient supply organisation. The interaction between stakeholders along the mill supply chain is
represented in two stages: the entire mill supply area and the logistic chains. Two models, based
on simulation rather than optimisation, have been designed for this purpose. The strategic model,
based on a simplified picture of the entire supply chain area, assesses the impact of stakeholders’
technical interfaces on total sugar production. The logistic model allows assessment of the
feasibility of strategic scenarios in terms of task capacities and daily management of cane flows.

This modelling approach also provides a valuable basis for informed discussion between
stakeholders. For example, new delivery allocation arrangements may lead to an alternative cane
payment system to maintain equity between growers. Grouping growers around common
equipment or encouraging contracting arrangements may lead to co-ordination problems between
growers and contractors, which will impact on cane flow regularity and cane quality (Le Gal and
Requis, 2002).
References:

1. Arjona E, Bueno G and Salazar L (2001). An activity simulation model for the analysis of the
harvesting and transportation systems of a sugarcane plantation. Computer and Electronics in
Agriculture 32: 247-264.

2. Barnes AJ, Meyer E, Hansen AC, de la Harpe ER and Lyne PW (1998). Simulation modelling
of sugarcane harvesting and transport delays. Proc S Afr Sug Technol Ass 72: 18-23.

3. Barnes AJ, Meyer E and Schmidt E (2000). Evaluation of methods to reduce harvest-to-crush
delays using a simulation model. Proc S Afr Sug Technol Ass 74: 25-28.

4. Chandra P and Fisher ML (1994). Coordination of production and distribution planning. Eur J
Opn Res 72: 503-517.

5. Cock JH, Luna CA and Palma A (2000). The trade-off between total harvestable production
and concentration of the economically useful yield component: cane tonnage and sugar content.
Field Crops Res 67(3): 257-262.

6. de Groote X (1994). Flexibility and marketing/manufacturing coordination. Int J Production

7. Eliashberg J and Steinberg R (1987). Marketing-production decisions in an industrial channel

8. Franck SD and Henderson DR (1992). Transaction costs as determinants of vertical co-

9. Gaucher S (2002). Organisation de filières et politiques d’approvisionnement. Analyse

10. Giraud-Héraud E, Soler LG and Tanguy H (1999). Avoiding double-marginalization in