This paper extends knowledge of disaster impact on the telecommunications power infrastructure by discussing the effects of Hurricane Katrina based on an on-site survey conducted in October 2005 and on public sources. It includes observations about power infrastructure damage in wire-line and wireless networks. In general, the impact on centralized network elements was more severe than on the distributed portion of the grids. The main cause of outage was lack of power due to fuel supply disruptions, flooding and security issues. This work also describes the means used to restore telecommunications services and proposes ways to improve logistics, such as coordinating portable generator set deployment among different network operators and reducing genset fuel consumption by installing permanent photovoltaic systems at sites where long electric outages are likely. One long term solution is to use of distributed generation. It also discusses the consequences on telecom power technology and practices since the storm.

DC power systems can be made more reliable by considering the load as an important energy asset. Currently the ability to manage the total system is available only through a centralized controller, which limits flexibility, reconfigurability, and reliability. These limitations can be avoided while still providing system level coordination through the use of distributed controls based on local information. All elements of the power system including source, loads, and the network itself have influence, interaction, and coupling to all other elements. Therefore, it is necessary to model and control a microgrid as a system of systems that share some common aspects, such as voltage levels, but can operate independently. Using local information in the form of the bus voltage, these techniques do not rely on a centralized controller, which improves system reliability. However, it is important to design the microgrid in such a manner as to take advantage of the energy not just from the generation sources, but also the energy stored in the individual points-of-load as well.

A power system is a collection of individual components that compete for system resources. This paper presents a game theoretic approach to the control decision process of individual sources and loads in small-scale and dc power systems. Framing the power system as a game between players facilitates the definition of individual objectives, which adds modularity and adaptability. The proposed methodology enhances the reliability and robustness of the system by avoiding the need for a central or supervisory control. It is also a way to integrate and combine supply and demand side management into a single approach. Examples are presented that use a simple nine bus dc power system to demonstrate the proposed method for various scenarios and player formulations.

Advanced metering infrastructure (AMI) deployment has been widely promoted in recent years to improve the accuracy of billing information as well as to facilitate implementation of demand response. Information integrity and availability of the devices is crucial to the billing information that should reflect accurately on how much the household energy is consumed. The IP-based smart metering devices may exist with unknown vulnerabilities that can introduce backdoors to enable worm propagation across AMI network. The infected devices can be attack agents that would largely disable the metering functionalities or manipulate control variables of each meter. This paper proposes an optimal frequency of on-site investigation and the number of monitoring verification to investigate potential anomalies of malware footprinting by applying the decision process framework of Markovian. The proposed method determines the best inspection strategies based on the observation from the existing anomaly detectors deployed in the network. The considerations include malware propagation characteristics, accuracy of anomaly detectors, and investigation and diagnosis costs. Four scenarios are simulated using the proposed method, demonstrating the effectiveness of investigation on potentially infected electronic meters within an AMI network.

Modern power electronics are capable of regulating loads with bandwidths so high that they essentially enforce constant power on millisecond timescales and contribute to system-wide voltage instability problems. Active front-end control of such loads that implement a power buffer function has been shown to mitigate instability, but has relied on complicated hybrid control techniques. This paper proposes a geometric control surface that implements the power buffer function by coupling the input impedance to the stored energy and by altering the source and load dynamics. The surface is derived from optimal control theory where importance is placed on maintaining continuous input impedance and retaining as much local energy as possible. The optimal control is a tradeoff between the needs of the system and the needs of the load. This paper introduces a geometric control surface based on a change of variables that simply and effectively implements a power buffer function. The formulation and implementation of the optimal surface are presented, in addition to experimental validation of the new power buffer control law.