Determining the Optimal Reserve Capacity of a Microgrid Based on the Probabilistic Analysis of Grid-connection in Market Environment

Abstract

Microgrid can be a useful entity to support stable and efficient operation of power systems with large-scale penetration of distributed generators, such as wind generators, energy storage systems, and combined heat and power plant. One major characteristics of microgrid is that it could take an island operation and maintain its reliable power supply if an accident occurred in the main grid. However, microgrid operator (MGO) cannot help taking some special action like load shedding during the island operation, since its generation capability has a limit. Therefore, MGO has to take this island operation into account when it make a plan for its own energy resources. Actually many prior researches about microgrid operation include reserve power scheduling in preparation for the uncertain islanding event. This dissertation analyses the risk of microgrid island operation, and describes the method that enables MGO to reflect this probabilistically into its operating cost when it makes a plan for the energy resources. In order to this, an islanding event of the microgrid is interpreted as a transaction suspension, and microgrid islanding rule is defined in the form of market rule to clarify the responsibility distribution of a contract breach in a market. To quantitatively examine the influence of market rule, different two microgrid islanding rules are proposed based on the Power Exchange for Frequency Control (PXFC) market, which was devised by M. Ilic et al. Postulating these two rules, the risk of microgrid island operation is examined. In other words, the triggering condition of islanding event is mathematically formulated, and microgrid islanding probability (MIP), which represents the probability of being in the islanded state during a unit time, is proposed and calculated. Utilizing the proposed MIP index, an optimization problem is constructed. The objective function is expected value of daily operating cost of microgrid, which include the risk of microgrid island operation, and the decision variable is the purchase capacity of reserve band in PXFC market. The optimization problem is solved and simulated with the market information of the PJM electricity market. The effectiveness of the proposed reserve scheduling method in terms of operating cost is investigated using simulations, where the proposed method and two further methods are applied to microgrids with different generation capabilities. Also simulation results of MIP analysis show that microgrid island operation has some hysteresis characteristics. Utilizing the proposed method, MGO can schedule its reserve power corresponding the market and grid conditions.