Development Of Decısıon Support Methods For Restoratıon In Earthquake-Damaged Medıum Voltage Electrıcıty Dıstrıbutıon Networks

Abstract

From past to present, there have been many destructive earthquakes in Turkey, they still do, these earthquakes show that this region is an earthquake region that has been shaken by serious damages and destructions. The 2020 Elazig and Izmir earthquakes have again revealed that this reality has always existed and how much damage it can cause; it is highly probable that the outcome of the Istanbul earthquake, which may take place at the same magnitude, will be much worse. An earthquake cannot be stopped or prevented, but the material and moral damage it causes can be minimized thanks to the preparations that can be made previously. Current estimations indicate that if another major earthquake occurs, significant damage will occur in earthquake-prone areas. One of the biggest problems that prevent the management of the post-earthquake process, which increases the loss of life after the earthquake, is the power cuts. Today, electricity has become the main source for the continuation of life and its constant presence in human life has thrown its importance in many areas out of focus, but the power cuts experienced in post-disaster situations have revealed that electricity can stop the activity of many things that work directly or indirectly today. Although transportation and communication are the biggest and main situations that it affects, the fact that it affects the operability of hospitals after the earthquake seriously triggers the loss of life. Therefore, after the earthquake, it is necessary to re-electrify as soon as possible. The focus of this study is to provide input data that can be used in the Markov Decision Process (MDP) based decision support system developed for re-electrification. Being able to accurately determine the post-earthquake condition of the electrical distribution systems is one of the basics for the correct operation of this system; in this study, a series of fragility analyses were performed to obtain the probability of damage to the electrical distribution systems. The main focus of the study is the buildings with electricity distribution systems in the basement in Istanbul, Kadıköy - Güzelyalı district. These electrical distribution systems not only supply electricity to the building under which they are located, but also distribute electricity to the surrounding buildings. After an earthquake that severely damages these buildings, said electrical distribution systems may become unusable and cannot distribute electricity to their surroundings. Some of these electricity distribution systems are located in locations critical to the entire electrical distribution system of the region, and severe damage to these stations can cause disruptions in the distribution system. To investigate this situation, possible damage to 6 specific buildings and their underlying electrical distribution systems was examined. As part of developing the decision support framework for the restoration of the medium voltage electricity distribution network damaged by the earthquake, detailed analyses of these buildings were carried out for the most critical locations in the said region. The buildings were modeled on real floor plans using the SAP2000 program, and nonlinear time history analysis was performed for 60 real earthquake records in each building. The analysis results obtained from the examined buildings were used to create a fragility curve for various damage situations. These fragility curves were then used as input for the Markov Decision Process (MDP) based decision support system developed to rapidly re-electrify the regional electricity distribution system.