Evaluating Energy Efficiency in Turkish Electric Distribution Using Network DEA and GA Models

Abstract

This study uses DEA to evaluate the energy efficiency of Turkey's provinces and electricity distribution companies. Then, the efficiency of electricity distribution companies is evaluated using network DEA by considering the sub-processes of electricity distribution companies, which consist of generation, transmission, and distribution. Finally, genetic algorithms were used to evaluate the efficiency of electricity generated from renewable energy sources in Turkey between 2006 and 2015 and compare them with traditional DEA. The GA used an initial population of 500 solutions and iterated for 1000 generations to determine the optimal input and output configurations for each distribution company. For 2015, only Bo & gbreve;azi & ccedil;i EDC is efficient among the twenty-one distribution companies evaluated in detail. The NDEA model with sub-processes provided more realistic efficiency scores than traditional DEA. Sakarya EDC performed outstandingly by achieving efficiency scores in all three processes: generation, transmission and distribution. This shows that Sakarya EDC operates optimally and is a benchmark for other companies in the sector. In particular, GA has demonstrated its effectiveness in navigating complex optimisation environments by providing efficiency scores that are, on average, 10% higher than those obtained with traditional DEA methods. Overall, the findings show that while NDEA provides a comprehensive energy efficiency analysis by breaking down processes, GA complements this by providing actionable insights for optimisation. This study provides valuable insights for policymakers in Turkey by emphasising the need for targeted investments in energy distribution to improve overall efficiency and sustainability.