Omya Hustadmarmor optimizes its Supply Chain

Table of Contents

I List of Figures

II List of Abbreviations

1 Introduction
1.1 Relevance
1.2 Research Objective and Outline

2 About Omya Hustadmarmor
2.1 Company Information
2.2 Supply Chain Management

3 Overcoming the Decision Difficulty
3.1 Operational Challenges
3.1.1 Complexity Drivers in the Supply Chain
3.1.2 Coherence of Production and Distribution Planning
3.2 The Inventory-Routing-Optimization Problem
3.2.1 Mixed-Integer Linear Programming
3.2.2 A Nonlinear Objective Function
3.2.3 Finding the Solution with a Metaheuristic Method
3.3 The Decision Support System
3.3.1 Design and Implementation
3.3.2 Beneficial Effects

4 Management Implications
4.1 Innovativeness of the Approach
4.2 Transferability to Other Industries

5 Conclusion and Outlook

III Appendices
Appendix A: The Efficiency of Single-Destination Shipping
Appendix B: Population-Based Genetic Algorithm
Appendix C: Literature on inventory routing problems in a marine context
Appendix D: Supply chain of Omya Hustadmarmor
Appendix E: Hustadmarmor’s yearly production output development

IV References

I List of Figures

Figure 5-1: Cost savings and quantitative benefits of the DSS implementation

II List of Abbreviations

illustration not visible in this excerpt

1 Introduction

Today’s business environment is characterized by an increasing complexity in manufacturing and supply chain management (Booth 1996). Huge numbers of choices and persistent time and margin pressures make managerial decisions more difficult. In addition, new enterprise applications and software are generating overwhelming amounts of data. Turning this data into decision-supportive information by applying theoretical models in practice is one of the tasks of Operations Research (Robinson 2000).

1.1 Relevance

For almost 35 years, the ‘Institute for Operations Research and Management Sciences’ (INFORMS) honors extraordinary practical achievements and success stories in Operations Research with the ‘Franz Edelman Award’. By bringing together top examples of innovation from different international organizations, the possibilities of Operations Research should be demonstrated and serve as a continuous source for new ideas and optimization. Among the finalists of the 2006 ‘Franz Edelman Competition’ is the European company Omya Hustadmarmor.

As a fast growing supplier for the European paper industry, Hustadmarmor faced an increasingly complex supply chain and decided to look for operations research-based planning support. Together with the Molde University College of Norway, the company conducted a project that led to the development of an innovative Decision Support System for maritime inventory routing (Informs 2006). This model allows significantly faster and superior management decisions and increases the predictability and, thus, the flexibility of Hustadmarmor’s supply chain.

1.2 Research Objective and Outline

Omya Hustadmarmor has presented a successful and innovative model for its individual supply chain challenges. In this paper the approach of the company to overcome the increasing supply chain complexity is analyzed. Moreover, further beneficial application of the model to alternative businesses will be discussed.

Following this introduction, the company and its supply chain will be introduced in chapter two. Chapter three describes the underlying inventory-routing-problem in detail. Firstly, the problem will be discussed on a theoretical basis. Secondly, the practical implementation of a Decision Support System to solve the problem will be discussed.

In chapter four the innovativeness of the approach is presented by a short literature review, followed by a discussion of a transfer of the applied method to different industries and possible extensions of the model. Chapter five concludes this paper with summarizing the key findings and benefits of the project.

2 About Omya Hustadmarmor

The research in this paper is examined from the empirical perspective of Omya Hustadmarmor. Consequently, the following section provides basic information about the company, followed by a description of the company’s supply chain management.

2.1 Company Information

The Swiss-based Omya Group is an international white minerals company supplying high- quality calcium carbonates and talcs to European paper manufacturers. The group employees around 5,600 people in over 30 countries, serving major markets including paper, plastics, rubber, coatings, adhesives, building products, and agriculture (Omya 2006).

Within the Omya Group, the Norwegian company Hustadmarmor AS is the largest production unit. Based near Molde, a small town located at the western coast of Norway, Hustadmarmor supplies the European paper-making industry with essential mineral additives, e.g. calcium carbonate.¹ Since its establishment in 1978 as a joint venture between Omya and the local Hustadkalk AS, the company has grown steadily and is today the world’s largest producer of pigments for paper manufacturers (Informs 2006). Starting with a production volume of less than 200,000 metric tons in the early 1980s, the company has invested repeatedly in production and storage capacity leading to the current output rate of over three million metric tons per year (see Appendix E). The downside of this steep growth, however, was a significantly increasing complexity in the company’s planning procedures. Since 99% of the plant’s yearly production is exported to other European countries, Omya Hustadmarmor faced a situation with overwhelming operational challenges in its production and outbound logistics.

2.2 Supply Chain Management

A supply chain² is essentially a network of firms engaged in manufacturing and assembly of parts to create a finished product (Tan 2000). In the context of Omya Hustadmarmor, the company functions as a supplier for the paper-making industry. Hustadmarmor transforms marble stone, which is delivered from quarries near the production site, into calcium carbonate slurry in a wet grinding process by adding chemicals and water.

Following the production procedure, the company is responsible for delivering 15 to 16 variants of the slurry to a number of storage tank farms. Altogether, there are 10 first-tier tank farms located at the coastlines of Germany, UK, Sweden, Netherlands and Finland. There is also a small number of second-tier tank farms near major customers, which however are of no relevance in this paper. In the farm tanks, the slurry is stored until there is demand from the local paper mills, in which the final paper products are manufactured (see Appendix D).

The physical environment (Hesse and Rodrigue 2004) determines which modes of transport can be used for transporting goods. Since both supplier and first-tank farms are located close to the coast, the calcium carbonate slurry is transported by ship. Together with the third-party shipping company Anders Utkilens Rederi, Omya Hustadmarmor owns a fleet of tank vessels of different sizes.³ The company uses direct shipping, i.e. only single-destination routes instead of a combination of multiple deliveries, since this method has proven as most efficient (Gallego and Simchi-Levi 1990, see Appendix A).

3 Overcoming the Decision Difficulty

This chapter is divided in three parts. In the first section the operational challenges of Omya are summarized. Based on this, the decision problem is analyzed theoretically in the second section. In the third section the measures applied in practice to overcome the challenges are described, i.e. the development of a decision support system.

3.1 Operational Challenges

3.1.1 Complexity Drivers in the Supply Chain

Omya’s supply chain challenges are primarily based on planning the three interrelated components, i.e. transportation planning, inventory planning and production planning.

With the company’s expansion two transportation planning elements have become increasingly complex: Firstly, the schedule of vessels, i.e. which vessel should depart on which day with which destination. Secondly, the shipment quantities, i.e. how much slurry and which product mix of the 15-16 variants should be transported. By addressing these two issues, planners at the company can ensure a constant supply of calcium carbonate slurry.

Since stock-out situations at the tank farms would lead to high follow-up costs at the paper mills, the inventory level need to be kept continuously above the safety stock as a necessary hedge against the underlying uncertain demand (Cachon and Terwiesch 2004). Demand of the paper mills for different slurry variants is very unstable, hence the safety stocks at the tank farms are traditionally relatively high. As a negative result, however, this implicates limited space left for replenishment stock at the tank farms and, thus, limits the use of large vessels. Using large vessels would allow economies of scale⁴, especially since marine transportation costs are very high and contribute a major part of the total costs: “The daily operating costs of a ship can be tens of thousands of dollars” (Christiansen et al. 2004, p. 2). Hence, improving fleet utilization could be translated into considerable improvements in financial results.

Also, it is important to find a proper vessel utilization because free vessels are usually used in the spot market (e.g. to transport methanol across the Baltic Sea). Hence opportunity costs are significantly high, and an inefficient vessel schedule would limit potential revenues from the spot market.

As a consequence of the vessel transportation, delivery times are long and imprecise. Transportation from the plant to the first-tier tank farms requires usually between two to five days in summer and six days in winter. External effects like the weather or mechanical defects may also influence the delivery time negatively (Ronen 1983).

From an inventory planning perspective, it is important that the company’s production and storage capacities at the plant are limited. The choice of a certain product mix for one vessel could eventually restrict following deliveries of this product mix. This shortage could inhibit using the full capacity of ships or prevent a timely replenishment of the tank farms.

Finally, in production planning it has to be considered that not all slurry variants can be produced at the same time, and that the production procedures for each of the 16 slurry variants need to be adjusted individually. This changeover time can be significant. Hence minimizing changeovers through producing preferably large production lots allow much higher production efficiency.

3.1.2 Coherence of Production and Distribution Planning

Omya’s management is challenged to thoroughly coordinate production and distribution planning along the supply chain. Before the project, the schedule planning was an informal process involving a number of parties at different stages of the supply chain. Basically, it can be seen as a manual demand-driven replenishment system: Whenever inventory at a local tank farm reached a critical level, the centralized Omya Sales Office was informed and ordered a vessel with the appropriate quantity at the production plant. To overcome the multiple operational challenges and maintain the needed level of operational flexibility, Omya traditionally has been using primarily small vessels.

However, increasing expenditures for electricity, fuel, raw materials etc. created a financial burden for the company and required a more cost-efficient planning. Hence, Hustadmarmor decided to use an increasing number of large vessels. As described, large vessels would allow the exploitation of significant economies of scale; but simultaneously they reduce flexibility of the supply chain and especially the predictability of the schedule planning. Frequent changes in production and distribution plans caused a number of domino-effects and led not only to delayed replenishments but also to quality issues and time consuming reprocesses.

3.2 The Inventory-Routing-Optimization Problem

Developing a planning tool for vessel transportation was seen as an adequate means to stabilize the company’s whole supply chain. This planning would need to include both of the main issues, namely routing vessels and replenishing inventory at the tank farms. Thus, the problem can be defined as an Inventory-Routing Problem (IRP) (Kleywegt et al. 2004). The overall objective is to find a cost efficient distribution plan, i.e. schedules, tank farms and product quantities for each vessel within a certain planning horizon, by not violating any of the given capacity constraints, i.e. neither stopping production at the plant, nor running out of stock at the tank farms.

3.2.1 Mixed-Integer Linear Programming

In a first approach, the optimization problem was modeled as a linear mixed-integer program (MIP), i.e. the minimization (or maximization) of a linear objective function subject to linear constraints with partly integer variables (Silver et al. 1998). The objective function (1) was defined as minimizing the overall distribution cost, concretely the sum of transportation cost and inventory cost,

Abbildung in dieser Leseprobe nicht enthalten

with P: number of products, J: number of vessels, K: number of tank farms, T: number of days in the horizon, c jk: total cost for sending vessel j to tank farm k, h pk: inventory cost for product p at tank farm k per product unit and per day, I pkt: inventory level of product p at tank farm k at the end of day t, Y jkt: {0, 1} variable that indicates whether vessel j leaves from the factory for tank farm k in day t (t Ç T − a jk), a jk: number of days for vessel j to go to tank farm k.

However, in this first approach, the linear mixed-integer linear programming could not deliver optimal solutions. This was due to the high number of uncertainties and operational challenges interrelated with the planning procedures.

[...]

¹ Today’s paper manufacture use an increasing share of minerals in the paper creating process, since mineral-enriched papers show better results in print performance and quality, and are an effective replacement of costly materials based mainly on wood or fibers.

² A supply chain is defined as a system of material suppliers, production facilities, distribution services and customers linked together via a forward flow of materials and a backward flow of information (Christopher 1998; Stevens 1989).

³ Vessel capacity ranges from 2,400 to 16,000 metric tons.

⁴ Here, economies of scale relate to the significantly lower cost per ton of large vessels in comparison to small vessels.