Logo image
Decision for relative weightings of the upstream-downstream layers of solar supply chain
Thesis

Decision for relative weightings of the upstream-downstream layers of solar supply chain

黃詠舜
Masters, 國立清華大學, 工業工程與工程管理學系
2011

Abstract

太陽能 分析網路程序法 供應鏈管理 決策分析 Solar Industry Analytical Network Process Supply Chain Management Decision Analysis
To survive in the highly competitive and volatile solar business environment, the ability to decide what layers along the supply chain to invest in and develop internally and which to leave others to develop is a critical problem for players in solar industry. For example, the supply shortage of polysilicon in 2005 severely tested the manufacturers in the industry. Long-term contract and vertical integration was the strategy to acquire raw material. However, when the financial crisis in 2008 made governments cut off the subsidies, the demand dropped dramatically. The price of raw material fell down rapidly and the contract and integration become a huge burden. We can observe that both make and buy is a strategy for solar manufacturers to maintain the competitiveness. However, only few researches propose comprehensive make or buy methodology, and concurrently make and buy was not an option. This research aims to build relative weightings of upstream-downstream layers decision framework for solar industry based on analytical network process (ANP), to help decision maker to decide an appropriate management structure in a supply chain. The tasks involved in this research includes: (1) Reviewing and sorting out critical factors for integration in economics theories and empirical studies. (2) Taking dependences and feedback in criterion and alternatives into the decision framework. (3) Derived the relative weightings between layers through ANP pairwise comparison, taking into account decision maker’s preferences and judgments. (4) Empirical study included face to face surveys with three solar companies in Taiwan, and the results were analyzed and discussed, including a sensitivity analysis.

Metrics

1 Record Views

Details

Logo image