台電輸電系統之生產力與效率分析

Abstract

本研究採兩階段方法，分析民國89-93年期間台電輸電系統之生產力與效率，第一階段應用資料包絡分析法（data envelopment analysis;DEA），依投入項不同的類別建立成本、設備、品質及總體模型，衡量六個供電區營運處之相對效率。第二階段進行Tobit迴歸分析，探討經營無效率的影響因素。最後再以Malmquist生產力指數，分析歷年來生產力的變動情形及累積成長率。 茲將本研究主要實證結果彙總如下： 1、在設備模型的效率分析中，供電轄區位於台灣中、南部的供電區營運處屬於無效率的單位，表示台電第六輸變電工程的內容應做適當的調整，以避免中、南部因過度投資導致輸電設備利用率低，造成資源浪費的情形。 2、由成本、設備及品質模型等不同構面的效率分析結果發現，在單一模型中具有效率的單位，在另一個模型中可能並不具有效率。 3、在敏感度分析方面，就整個輸電系統而言，去除變壓器裝置容量投入項後，效率值變動最大，故變壓器裝置容量是影響效率的關鍵因素，其次則是事故次數。 4、在Malmquist生產力指數分析方面，六區的總平均累積成長率達7.15％，表示整個輸電系統營運單位的總要素生產力是處於成長的情形。而生產力的成長主要是源自於技術的成長。 5、在Tobit迴歸分析方面，人口密度及特高壓大用戶占全台比率越高的地區，越有助於提升輸電系統的效率。而輸電線路下地率越高，越有助於品質模型的效率提升，但另一方面也將造成成本模型的效率降低。

This paper applies a two-stage method to examine the productivity and efficiency of transmission systems of the Taiwan Power Company (TPC) during the 2000-2004 period. In the first stage data envelopment analysis (DEA) is used to measure the relative efficiency of six electricity transmission districts. This study constructs four DEA models, according to the input categories of cost, equipment, quality, and all of the above. In the second stage a Tobit regression is used to find the effects of environmental variables on efficiency scores. Finally, the Malmquist productivity index is used to compute the longitudinal total-factor productivity (TFP) changes. Our major findings are as follows: 1. Some electricity transmission districts that perform well in one model may not perform well in another model. 2. The population density and rate of extra-high voltage customers have positive impacts on the efficiency of a transmission system. 3. The rate of underground cables has a positive impact on efficiency in the quality model but a negative impact on efficiency in the cost model. 4. Their average accumulative growth rate of TFP is up to 7.15%. 5. Their TFP growth is mainly due to technology innovation.