Implementation of Distributed Wide Area Monitoring System Server Configuration on Java-Madura-Bali Power System

PLN, as the Indonesian electricity company, has implemented large-scale monitoring systems. One of these monitoring systems is Wide Area Monitoring Systems (WAMS), which aims to monitor any measurements in the Java-Madura-Bali (Jamali) power system. The WAMS server in the control center and the phasor measurement unit (PMU) in the substation connect to each other. The centralized WAMS server has two functions: phasor data concentrator (PDC) and historian. Initially, the WAMS on the Java-Madura-Bali system could monitor as many as 38 PMUs from various locations, with a CPU utilization of 57%, and the WAMS server could operate normally. When 32 PMUs were added, problems arose because the CPU utilization on the WAMS server reached 100%. This happens because one server does all the processing for the PDC, historian, and vector processor, so all data in WAMS is lost. The test results show that the historian requires more CPU and memory resources when compared to the PDC. Thus, the PDC and historian applications need to be separated on different servers. Based on previous research related to distributed real-time servers, this paper provides a short-term solution in the form of separating PDC and historian servers and a long-term solution for the Java-Madura-Bali WAMS configuration. The short-term solution results in the performance status on the historian server being at 8%, so WAMS is in a safe condition. The long-term solution is in the form of a regionalization system where PDC servers are located in each area or region. This solution can be used as a reference for other power systems that will implement WAMS.