在人類文明的快速發展下,人類大量使用石化能源和排放二氧化碳等溫室氣體,使得石油含量即將耗盡、臭氧層破洞及氣候暖化,而近年全球因為氣候異常,災害頻傳。世界各國為解決這些問題,便開始規畫並推動相關政策,其中,再生能源的發展是各國政府所推動的重點。再生能源是利用大自然的資源發電,例如:風力、太陽能、潮汐、地熱等,不但不會排碳,也有用之不竭的特性,因此若能以再生能源替代傳統石化能源,不但能減少對於石油的需求、降低二氧化碳的排放,更能達到永續的效果。台灣也跟著國際的減碳潮流,陸續推動再生能源發展條例、永續能源政策綱領等政策,期望可以增加再生能源的使用。而躉購費率制度是台灣政府透過較高的固定電價,保證收購再生能源的發電量,以鼓勵民間投入再生能源的方式。每年政府會依照再生能源的建置成本、年運轉維護費、年售電量等因子做躉購費率的計算,若是費率訂得太低,便無法吸引民間投資再生能源;訂得太高,則會對於政府財政造成龐大的負擔。因此,若能對於再生能源之建置成本有較完善的了解,便可以進一步研究躉購費率的變化,故本研究即透過再生能源建置成本學習曲線的建立,以分析其建置成本的未來發展趨勢,然而一般學習曲線僅能夠描繪裝置容量與建置成本的關係,本研究提出一階層學習曲線之模式,除裝置容量和建置成本外,還考慮其他變數,並將此模型應用於台灣風力發電和太陽能發電。研究結果顯示,階層學習曲線較一般學習曲線有更佳的預測值,故期望後續研究可將此階層模型應用於其他領域中。
Due to the industrialization, people consume large amounts of fossil fuels and emit lots of carbon dioxide. Experts predict that fossil fuels are going to use up and the temperature of earth increases yearly. The climate change even results in some serious disasters. Therefore, countries execute policies to solve the problems of the demand of fossil fuels and the climate change. Because renewable energy comes from natural resources and will never deplete, governments put the emphasis on the development of renewable energy. Taiwan government also follows the trend of carbon reduction and passes the Statute for Renewable Energy Development. The statute guarantees that the government will buy back the electricity generated from renewable energy with a fixed price in twenty years. Therefore, if the price is set too low, it is unattractive to public; on the contrary, if the price is set too high, it is a huge financial burden for the government. The price is calculated yearly based on the installation costs. Thus, by studying the installation costs can help understand the buy-back price. This research uses a learning curve model to predict the change of the installation costs in the future. However, studies usually study the costs by one factor learning curve model, which cannot thoroughly depict the relationship between the installation costs and the cumulative capacity. Therefore, the research proposes a hierarchical learning curve model and uses Taiwan’s wind power and photovoltaic as case study. The results show that the hierarchical model predicts the installation costs more accurately than the basic one factor learning model. The hierarchical model can also be applied to other kinds of renewable energy and even other research area.