New Solutions for Optimization of the Electrical Distribution System Availability in Microgrids: Application to Petroleum Platform in Tunisia Mohamed Ghaieth Abidi * , Moncef Ben Smida † , and Mohamed Khalgui *‡§ * LISI Lab, INSAT Institute, University of Carthage, Tunisia. Email: Mohamed.ghaieth@gmail.com † LSA Lab, Tunisia Polytechnic School, University of Carthage, Tunisia ‡ School of Electrical and Information Engineering, Jinan University (Zhuhai Campus), Zhuhai 519070, China. § SystemsControl Lab, School of Electro-Mechanical Engineering, Xidian University, Xi’an 710071, China. Abstract—This paper deals with the availability in microgrids that are composed of a set of sources (Photovoltaic generators, wind turbines, diesel generators and batteries) and a set of loads (critical and uncritical loads). The energy produced by various sources will be grouped in an alternative bus (AC bus), and it will be distributed on loads through an electrical distribution system. The occurrence of a fault in the system can cause a total or partial unavailability of energy required by the loads. The objective of this paper is to characterize the fault caused by the limited reliability of the components of the electrical distribution system and to propose an new design methodology to optimize the availability of this system (as well as the availability of power supply) by taking into account all the economic constraints. The proposed methodology is based on the redundancy of electrical distribution paths. An application of this optimization to a petroleum platform shows clearly a high degree of supply availability distribution in microgrid. Index Terms—Microgrid; Electrical distribution system; Re- liability; Reconfiguration; Fault tolerance; High power supply availability. I. I NTRODUCTION For various reasons, conventional power systems are con- fronted with economical and technical problems in terms of quality of electrical power distribution services. These problems have led to a new trend, relying on local production of energy by using non-conventional sources of energy (dis- tributed sources) [1]. Microgrid system is a new concept that aims to integrate decentralized energy sources efficiently and reliably [3]. A microgrid is composed of sources (Photovoltaic cells, Wind turbine, Battery, Diesel generator. . . ) and loads. The energy produced by diverse sources will be grouped in a bus (AC or DC bus), and it will be distributed on loads thanks to an electrical distribution system. This system consists of switchgears connected between them by electric lines. The components of a distribution system can have breakdowns. These failures (breakdowns) can result a total or partial unavailability of energy required by the loads. For many sector applications, the power supply availability are regarded as paramount factors. Power unavailability can cause material and human losses. One of the major advantages of microgrid concept is that the transmission costs are reduced and long distance energy transfers are no more necessary. This scale reduction of the grid gave us more flexibility in terms of design (configuration and reconfiguration) of power systems, essentially concerning the distribution system level. In order to improve the quality of service and to have a high availability of energy, it is necessary to design the power system that has an optimal architecture and size. The chosen architecture must ensure the transfer of the energy produced to the loads (especially the critical loads). Until now, the conception of a microgrid presents several hard problems [4]. In order to optimize the architecture, the microgrid will be modeled by a directed graph. Each node represents a switchgear and each connection will be presented as an edge. This representation should take into account the reliability of components as well as connections and minimize the effect of probable faults. The reduced size gives more flexibility at the level of the interconnections between the various components. The strategy of optimization has to use a probability analysis based on the rates of failure to calculate the rate of availability of the energy in nodes. After identifying the nodes with low availability, the strategy of optimization has to improve the availability by modifying the interconnections between the various nodes of the distribution system. The proposed solution is based on the multiplication of the paths of electrical power flow [13]. By adding new connections, new paths will be in passive redundancy with the existing paths. These new connections will be used in case of a defect within the initial connections in order to answer availability requirements of the electrical energy. These new connections will be an algorithm that promotes the low-cost connections with the nodes that have a high availability. Such a solution should take into account the modifications cost. In case of several possible solu- tions, the least expensive solution is chosen. This optimization strategy will be applied to a Tunisian petroleum platform to confirm its efficiency. This paper is organized as follows: Section 1 presents the state of the art of microgrid design optimization. The second Section explains the formalization of the distribution system. Section 3 proposes the Decision aid design solution for optimal availability. In Section 4, we