Mobile Edition
Print
Get the Mag
Weekly UpdatesAbstract
Kinmen Kaoliang Liquor Inc. (KKL), once a nonprofit government monopoly, was corporatized in 1998. Cost reduction and profit maximization became the new corporation's main goal. Located on an island off the main island of Taiwan, KKL sells over 90 percent of its produced liquor domestically. In order to restructure its distribution system, KKL is planning to establish the distribution centers (DCs) on the main island of Taiwan. This study aimed to provide the decision support from a quantitative aspect for KKL's strategic plan. The classic fixed-charge model was modified to take into consideration the inbound and outbound transportation costs of the DCs as well as the variable and fixed components of the DC facility costs. This study collected and estimated the relevant parameters in the model, and obtained the optimized solution. The results indicated that three DCs need to be established with a total system cost of NTD 110,786,400 per year.
**********
Kinmen Kaoliang Liquor Inc. (KKL) started out as a nonprofit government monopoly, and was corporatized in 1998. After the reform, cost reduction and profit maximization became the new corporation's main goal. Located on an island off the main island of Taiwan, KKL sells over 90 percent of the liquor it produces domestically. There are two major distributing channels: KKL itself (including three branch companies) and sales agents. KKL manages the logistics operations through third-party logistics service providers for the liquor they themselves sell. As to the other channel, the sales agents take care of the logistics operations from the plant in Kinmen. Although sales agents help to simplify the operations, KKL suffers from several drawbacks, such as a reduced profit margin and a rise in potential competitors. KKL has started to take back sales from the sales agents and is planning to establish the distribution centers (DCs) in charge of the logistics operations and trade transactions on the main island of Taiwan. These DCs may then become the forerunner of the marketing company that KKL plans to invest in for the next stage.
Although operations research techniques have been widely applied by companies to design or restructure their supply chain system (e.g., Laval et al. 2005 and Ulstein et al. 2006), most local companies in Taiwan make their decision to locate a DC somewhere from a qualitative point of view. Thus, this study aims to provide decision support using a quantitative approach for KKL's strategic plan to establish the DCs on the main island of Taiwan. Various quantitative methods have been used for locating DCs in several case studies. For example, Ehrgott and Rau (1999) developed general cost models for both cost and delivery time of a bi-criteria evaluation method to analyze the scenarios for improving the distribution network of a chemical company in Nordic countries. Nozick and Turnquist (2001) combined the fixed-charge model and the coverage model to consider transportation cost, inventory cost, and service level for locating the distribution centers of an auto manufacturing company in North America. Farahania and Asgari (2007) combined the MCDM (Multi-Criteria Decision Making) model and the covering model to determine the DCs in a military logistics system.
Given the unique features of KKL, this study modified the classic fixed-charge model to take into consideration the inbound and outbound transportation costs associated with the DCs as well as the variable and fixed components of the DC facility costs. This study collected the data from KKL to estimate the values of the parameters in the model and obtained the optimized solution. The result of the case study indicates that the total cost is NTD 110,786,400 per year (about USD 3,632,340), and that three DCs are to be established near Keelung, Changhua, and Kaohsiung in northern, central, and southern Taiwan respectively. According to the sensitivity analysis, the variations in fixed cost and demand level are the most influential factors for locating the DCs.
The remainder of this article is organized as follows: The next section provides the background of the company in the case study, followed by a section introducing the mathematical formulation, a modified fixed-charge model, as well as the approach and the result for estimating the parameters in the mixed integer programming (MIP) model. The numerical results from solving the MIP model are presented, followed by the sensitivity analysis and the scenario evaluation. Finally, the findings and conclusions of this study are summarized.
CASE BACKGROUND
Kinmen (also known as Quemoy), a small archipelago of several islands, is administered by the Kinmen County Government, a local government of Taiwan (ROC). The location of Kinmen is shown in Figure 1. Kinmen is geographically very near Xiamen, China (PRC) and is separated from the main island of Taiwan (also known as Formosa and later simply referred to as Taiwan) by the Taiwan Strait, which is on average about 200 kilometers wide. Due to its unique natural environment and outstanding manufacturing techniques, Kinmen has been famous for the production of kaoliang jiu, a strong distilled liquor made from fermented sorghum.
The first distillery of KKL was established in 1952 and re-named the Kinmen Distillery in 1956. Once a government monopoly, Kinmen Distillery was corporatized and became Kinmen Kaoliang Liquor Inc. in 1998. The liquor sales of KKL have been growing steadily. The revenue exceeded the milestone of 10 billion NTD in 2006 and reached 11.2 billion NTD (about 368 million USD) in 2007. Over 90 percent of the revenue comes from the domestic sales in Taiwan, although KKL has begun to pay more attention to foreign markets such as China and other countries.
Focusing on its core competence of liquor production, KKL traditionally has been relying on sales agents to expand its market and simplify its distribution operation. However, it also maintains its own distribution channels and three branch companies in Taiwan. A sales agent could become a competitor at the end of the agency contract if that agent turns to another liquor maker. In addition, the profit margin for KKL for sales made via agents tends to be low. Therefore, KKL has been reducing the amounts of sales it makes via its agents. As the nearly perfect quality of the kaoliang jiu produced by KKL depends heavily on the unique environment in Kinmen, it is unwise to move the production facility. However, KKL is planning to establish its own distribution centers in Taiwan to facilitate the distributing operation, thereby reducing the logistics cost and improving customer service.
Due to the geographical location, the delivery between the plant of KKL in Kinmen and the wholesalers or distributors in Taiwan involves several transportation links with multiple modes. The transportation network after the establishment of the DCs is illustrated in Figure 2.
The finished product is first transported from KKL's plant in Kinmen to the Liaolo Harbor (Link 1), the only port with sea transportation service in Kinmen to Taiwan. KKL owns and operates a private fleet of trucks for carrying raw material as well as finished products. Thus, this transportation link is well controlled, and the associated cost is relatively insignificant. As for the sea transportation between Kinmen and Taiwan (Link 2), several carriers are offering scheduled services between Liaolo and the three major seaports in Taiwan, Kaohsiung, Kee-lung, and Taichung. Even though the service to Taichung in central Taiwan has the shortest distance, it is the least frequent and the most expensive. The space on these cargo transporters may be limited or even unavailable during the peak season. In addition, sea transportation service is by nature relatively unreliable due to weather conditions. These factors make the Taiwan Strait link the most troublesome in terms of the whole logistics operation. At present, for products distributed by KKL, shipments are sent directly to the wholesalers from the ports by trucks. KKL currently has an outsourcing contract with several trucking companies in Taiwan. In the future, products will be sent to the DCs (Link 3) first to replenish the inventory. Once a wholesaler places an order, the shipment is assembled and delivered from the DC to the wholesaler (Link 4).
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
According to the arrangement of the transportation operations described above, the number and locations of the DC is critical to the overall transportation cost. Nevertheless, facility cost should also be taken into account in order to minimize the overall logistics cost. The model introduced in the next section is used to provide the decision support to balance the trade-off between transportation cost and facility cost. As to the inventory cost, KKL, a strong player with an overwhelming market share, usually does not maintain a high level of safety inventory, and thus inventory cost is not considered in the model.
MATHEMATIC PROGRAMMING MODEL AND PARAMETER ESTIMATION
Location analysis models have long been studied and presented in the literature. According to Daskin (1995), there are four common types of problem formulations: coverage problems, P-median problems, P-center problems, and fixed charge facility location problems (later referred to as fixed-charge models). For the models and the solution algorithms of location analysis, Mirchandani and Francis (1990) as well as Dresner (1995) serve as an excellent source.
Based on the background described in the previous section, this study chose the fixed-charge problem to model the decision problem. In the next subsection, the mathematical programming model is presented. In particular, the definitions of the parameters have been modified so that the classic model originally for two-staged problems can be applied to deal with the case with four transportation links, as illustrated in Figure 2. In the following subsection, the approach and the result for estimating the parameters in the mathematical programming model are described.
FEED