Akademik Veri Yönetim Sistemi
JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI, cilt.24, sa.1, ss.87-101, 2021 (ESCI)
In this study, performance measurement and efficiency benchmarking of Turkey's 21 electricity distribution companies were aimed. Random input and output variables of distribution companies are included in the stochastic input and output oriented chance constrained Banker, Charnes and Cooper (BCC) Data Envelopment Analysis (DEA) models within the symmetric error structure. Deterministic input and output oriented BCC DEA models were compared with the stochastic models and stochastic measurement values were verified with the deterministic measurement values. The chance-constrained problems were created by uncertain data for stochastic DEA, then stochastic models were converted to deterministic equivalents and were linearized. Turkey's 21 electricity distribution companies have been accepted as decision-making units (DMUs) of DEA models. According to the efficiency results of deterministic and stochastic models, efficient and inefficient electricity distribution companies were determined. The objective is to add statistical noise to DEA models in the analyzed data and to obtain performance measurement in this direction. Noise element was included into the models within the scope of single factor symmetric error structure and the models were examined separately for each error level in order to examine the different results between the models. This study is the first approach that was performed to evaluate the power distribution companies to incorporate stochastic data into DEA within a symmetric error structure in Turkey.