Sharma Kolluri

Decision Tree-Based Preventive and Corrective Control Applications for Dynamic Security Enhancement in Power Systems

In this paper, decision tree (DT)-based preventive and corrective control methods are proposed to enhance the dynamic security of power systems against the credible contingencies causing transient instabilities. Preventive and corrective controls such as generation rescheduling and load shedding schemes, respectively, are developed based on the security regions and boundaries that are calculated in the space of appropriate decision variables. The security regions and boundaries are determined by the rules of DTs that are developed by the generated knowledge bases. This work also involves improving the accuracy of security boundaries as well as the optimal solutions for the fuel cost and load shedding optimization problems encountered in the preventive and corrective controls. The methods are implemented on the Entergy power system model.

Decision tree assisted controlled islanding for preventing cascading events

At stressed operating conditions, critical contingencies can initiate loss of synchronism and trigger cascading events. Controlled islanding is the last line of defense to stabilize the whole system. This paper presents a decision tree assisted scheme to determine the timing of controlled islanding in real time by using phasor measurements. In addition, a slow coherency based approach is used to determine where to island. This scheme is tested on the operational model of the Entergy system and a severe N-2 outage case is used to demonstrate the phenomenon of cascading events due to protective relay actions. The results show that training one decision tree only for a specified critical contingency that can potentially cause cascading events can yields high prediction accuracy. Being aware of loss of synchronism in real time, operators can implement controlled islanding at carefully designed transmission interfaces and rapidly stabilize each island. Thus a significant amount of load is still served compared to uncontrolled system islanding.

Innovative approach for solving dynamic voltage stability problem on the Entergy System

The Western Region of the Entergy System is limited in both generation and transmission. Past events as well as extensive voltage stability analysis had indicated that the region could be subjected to severe voltage instability. In 1998, Entergy implemented a fast acting load shedding scheme called VSHED. However, recent studies have indicated that due to increased load growth and transmission limitations in the region, the voltage instability problem could be more dynamic in nature. In order to mitigate this problem, Entergy considered implementing several existing as well as new technologies such as SVC, STATCOM, and distributed superconducting magnetic energy storage (D-SMES). Based on extensive evaluation, D-SMES proved to be the best alternative. This paper discusses the technical assessment carried out to address the voltage stability issues as well as various alternatives considered in identifying the optimal solution.

Island formation in entergy power grid during Hurricane Gustav

Hurricane Gustav made landfall on September 1, 2008, Southwest of New Orleans, LA. Several transmission lines were outaged and an electrical island was formed in an area including the metropolitan New Orleans, LA. The island was sustained for several hours before the system was restored. This paper discusses the issues associated with maintaining an island, results of the dynamic simulations performed to replicate the island, and issues associated with restoring the island.

Power System Connectivity Monitoring Using a Graph Theory Network Flow Algorithm

A method of applying network flow analyses during real time power system operation, to provide better network connectivity visualization, is developed and presented. Graph theory network flow analysis is capable of determining the maximum flow that can be transported between two nodes within a directed graph. These network flow algorithms are applied to a graphical representation of a power system topology to determine the minimum number of system branches needed to be lost in order to guarantee disconnecting the two nodes in the system that are selected. The number of system branches that are found serves as an approximate indicator of system vulnerabilities. The method used in these connectivity analyses makes use of well known graph theory network flow maximum flow algorithms, but also introduces a new algorithm for updating an old network flow solution for the loss of only a single system branch. The proposed new algorithm allows for significantly decreased solution time that is desired in a real-time environment. The value of using the proposed method is illustrated by using a detailed example of the 2008 island formation that occurred in the Entergy power system. The method was applied to a recreation of the 2008 event using a 20,000-bus model of the Entergy system to show both the proposed methods benefits as well as practicality of implementation.

Voltage stability assessment of a large power system

Modal analysis and reactive power reserve method are applied in the Entergy electrical system to find the buses and zones prone to voltage instability. In the modal analysis method, weak zones are identified by monitoring the participation factor of the buses in the critical modes. In the reactive power reserve method, reactive power reserves of the zones are monitored for single line outages, generator outages, and double line outages and the zones that have exhausted their reactive power reserves are identified as critical. The critical contingencies resulting in the smallest stability margins are also computed, and are ranked in the order of severity. The voltage stability assessment tool (VSAT) developed by Powertech Labs, is used to determine the weak zones and to determine the critical contingencies.

Development of an Alternative Black- Start Plan for System Restoration

Following hurricanes Katrina and Rita that hit the Louisiana gulf coast, the designated black-start generating unit in southern Louisiana suffered extensive damage. Studies were performed to identify alternate means of restoring the system in the event of a system-wide black out. One of the options considered was to utilize an existing black-start unit in Northern Louisiana to energize the existing 500 kV lines to Southern Louisiana in order to provide start-up power for the generating units in that region. This paper discusses critical issues related to generator, substation and transmission equipment that were considered while developing the black-start plan

Planning and implementation of large synchronously switched shunt capacitor banks in the Entergy system

The Western Part of the Entergy system is one of the fastest growing in the system with a high concentration of industrial loads. The power into the area is imported primarily over Richard-Webre and Mt. Olive-Hartburg 500 kV lines. The loss of the Richard-Webre line coupled with an outage of the largest generator in the area (Nelson 6 unit) leads to a reactive power deficiency in the area resulting in a voltage collapse. Amongst the various solutions for the voltage collapse problem, the installation of large shunt capacitor banks was deemed to be one of the most cost-effective option and one that had the least implementation time. Therefore, 140 MVAr capacitor banks were recommended for the Nelson 230 kV and Carlyss 138 kV substations. These capacitor banks were the largest ever installed on the Entergy system. Due to the extremely sensitive nature of loads in the area and potential voltage magnification problems, it was decided to use synchronous closing breakers to switch these capacitor banks. Transient studies were performed to determine the impact of overvoltages introduced during switching of the capacitor banks. This paper focuses on factors that influenced bank sizing and location, results of the switching studies, choice of switching device, installation and operating experience.

Automated Critical Clearing Time calculation for analyzing faults at Entergy

This paper describes the implementation of Critical Clearing Time (CCT) software at Entergy for transient stability analysis. The software utilizes a fast time-domain simulation method to compute critical clearing time under different fault conditions. The framework presented in this paper allows for automated critical clearing time computations that eliminates manual time-consuming cut-and-try critical clearing time calculations. The paper discusses how automatic CCT computations are applied at Entergy for addressing various protection system concerns.

A Potential Island Formation Identification Scheme Supported by PMU Measurements

An island formation identification scheme utilizing both phasor measurement unit (PMU) measurements and supervisory control and data acquisition (SCADA) information, in conjunction with an analytical method, is developed and presented. The motivation for this work came from the investigation of the islanding event in the Entergy, Louisiana power system in 2008. During Hurricane Gustav in September of 2008, 14 transmission lines were lost which led to the formation of a large electrical island containing much of Baton Rouge and New Orleans. This work proposes a method for identifying any specific area of the transmission system that is vulnerable to islanding as the result of losing a system branch. PMU measurements are then utilized to verify SCADA reports that a system branch has been lost. The proposed identification scheme is called the one line remaining (OLR) algorithm. The proposed algorithm identifies potential island formations based on the criteria that a single bus, or a number of buses, are connected to the rest of the system by only a single branch as a direct result of a branch outage. The OLR algorithm is detailed and applied to several small test systems to motivate and validate the approach. The OLR algorithm is then applied to a large Entergy test system for the Gustav islanding event and correctly identifies the last tie line that was lost and created the island.