Power pooling among electric utility companies aims at effectively harnessing operating economics and reliability benefits through coordinated interchange of power, energy, and related services. In the existing utility industry structure, the operation of power pools brings about the necessary level of coordination to maintain the integrity of large interconnections. In light of growing competition, the continuation of such pools is difficult. This project examines the structures of existing and proposed power pools. It aims to construct analytical frameworks for such coordinated operations. The work will explore the minimal requirements for coordination to maintain system reliability and security. The frameworks will also be used to assess the economic efficiency of pooling.

The most challenging aspects of multiarea studies is to model realistically the loads and resources in each area and to construct computationally efficient schemes for their simulation. Typical applications are to interchange contract evaluation, geographically differentiated marginal costing studies, transmission services pricing, and strategic and resource planning. The multiarea production cost simulation model must correctly take into account the impacts of transmission constraints as well as interconnection operational policies. Our objective is to build a general model to simulate the operation of multiarea power systems under various operational policies, ranging from totally centralized dispatch to decentralized bidding dispatch.

We have developed a general framework for the analysis of competitive electricity markets modeled after the Power Pool of the England and Wales electric supply industry. Under the assumption of perfect competition, we derived the formulation of optimal bidding strategies. Using this framework and the analytical results to date, we are investigating the consideration of demand side bidding in the electricity markets. Strategies for maximizing profits of demand side bidders will be formulated. Additional areas of investigation are the relaxation of the perfect competition assumption, the study of market power concentration in the hands of few players, and the incorporation of financial contracts into the strategies of bidders.

Soon electric utilities will be required to implement real-time information networks giving all transmission customers equal access to all information on the availability of transmission capacity. These transmission service information networks (TSINs) will be required to convey large quantities of data among multiple users and to do so very reliably. The goal of this work is to analyze the system requirements of a TSIN and how an electric utility may build such a network, interface it with the networks of other utilities, and cooperate with other carriers to provide commercial telecommunications services. Economic and regulatory issues as well as the technical details of building such a network are being explored.

AIMS is a computerized hourly interchange matching system whose goal is to promote the maximum economic savings among all the participating players. This is accomplished by matching of bids to sell and offers to buy so that the sum of the savings for all the participants is maximized. We are evaluating the matching scheme from the point of view of the system, a buyer, and a seller. Our interest is to study the strategic behavior of players in formulating their bids to sell and offers to buy. We are investigating the truth revelation characteristics of the bids/offers, the role of transmission availability and the overall impact on system operations.

In a restructured environment, electric utility consumers will eventually choose providers of electrical energy. Hence, there will be greater use of the system for transmission between various players and a much higher level of power flowing through the power grid. This, in turn, will bring about the need to quantify the amount of transmission service that a network can provide. Our research aims to develop a consistent definition of transmission transfer capability and a general set of procedures for its evaluation. We will investigate the information requirements and the computational aspects and will study the use of a real-time information network as a medium for sharing the necessary information among various parties involved in the transmission of electricity.

Advances in digital and analog circuits are pushing power supply levels to 3 V, 2 V, and even lower. These low voltages help decrease power requirements in complex circuits. Efficient conversion of energy at voltages below conventional logic levels is very difficult. Circuit arrangements and control approaches suitable for levels as low as 1 V are being explored. Stable large-signal operating techniques are being developed to provide acceptable performance at low output levels. Synchronous rectifiers and related techniques are being studied for low-voltage applications. A flexible integrated power conversion system is being prepared for fabrication under a MOSIS CMOS process.

Power conversion circuits are large-signal nonlinear networks controlled exclusively through the action of switches. Several new approaches are being developed for power converter control. One approach explands on geometric methods, such as sliding mode control, used successfully in other nonlinear applications. In this boundary control approach, geometric structures in state space are used to control the evolution of converter voltages and currents. Methods such as boundary control offer precise, reliable converter operation with minimum influence by unknown parameters and external noise.

A complete hybrid electric car, combining an electric traction system with an engine-generator set, has been built and is now under study in the laboratory and on the highway. The car is designed to meet all performance, safety, and convenience characteristics of standard automobiles, while reducing exhaust emissions by as much as 90%. Objectives are to characterize major subsystems of a practical hybrid car in depth. Tests of efficiency, fuel economy, and emissions are being conducted. Parametric studies of subsystems are in progress. The data and information will assist industrial firms in the evaluation, design, and development of hybrid vehicle technology.

Electrohydrodynamics of a Conductive Liquid Meniscus
J. C. Chato,* P. T. Krein,* G. Wright
Tektronix, Inc.; Power Affiliates Program
(In conjunction with the Department of Mechanical and Industrial Engineering)

This program involves the control of liquid drops with electric fields. Objectives are to determine liquid properties important to the interactions, and to identify conditions under which a meniscus can be disrupted with a field. Possible applications include liquid spraying and electrically activated ink jets. Boundary element field modeling is in use for simulation studies, in conjunction with an experimental test bed.

Field orientation is a widely used control method for ac induction motors. Recent results in nonlinear control theory, including feedback linearization and integrator backstepping, offer possible alternatives for ac servo systems. Observer techniques allow high performance without expensive sensors. This project examines the operating performance of new motor control alternatives. Methods are studied analytically, through detailed simulation, and experimentally. A digital signal processing motor drive system has been designed and built for tests.

Comparisons are being made among various simulation approaches for switching power conversion systems. The switching nonlinearities of these systems are well suited to piecewise simulation approaches, but less well suited to conventional methods. The project compares SPICE-based circuit simulators and mathematical simulation methods such as MATLAB. The objective is to learn the considerations needed when preparing a simulation tool suitable for power electronics modeling and analysis.

Pulse-width-modulated inverters are experiencing growing application for control of ac motors. Modern systems support motors at power levels up to about 100 kW, although cost increases rapidly above 20 kW or so. An alternative at high power levels is to use several inverters in parallel. To make such an arrangement reliable, tight coordination of individual inverters is necessary. The project is studying coordination techniques. Both device-level and system-level approaches are being examined through analysis, simulation, and experimental tests.

The electric utility industry is entering one of the most turbulent periods of change in its entire history. The current pressures of regulation and competition, combined with the rapid influx of new information, communication, and control technologies, make it difficult to accurately predict the eventual structure of the industry. However, given the complexities of the power system, it is important that the industry undertake a managed redesign. To accomplish this, it is important to be able to communicate clearly the complex issues involved in power system operation and to identify the essential operating requirements needed to maintain security under any restructuring alternative. In this project we are examining the use of computer visualization/simulation to achieve these goals.

The Newton-Raphson power flow is used as a major tool for analysis of the nonlinear power system loading equations, though usually with moderately loaded systems without the need to consider a large number of device limitations. However, future systems will be more stressed, and as advanced electronic control devices proliferate there will be a vast increase in the number of device limits that must be considered. This research is investigating the convergence properties of the Newton-Raphson power flow for such cases, with the goal of developing improved solution methods.

As power systems become more heavily loaded, system operation will be increasingly constrained by contingent cases for which the power flow equations have no real solution. The goal of this project is to develop a measure to quantify the unsolvability of such cases and to determine the optimal controls to restore the case to solvability. A Euclidean norm is used in parameter space to measure the degree of unsolvability. The sensitivity of this measure to different system controls is then used to determine the best controls to restore the case to solvability. Both the static and dynamics aspects of the problem are considered.

In the restructuring of the electric power industry, a number of alternative paradigms for the future industry structure are under consideration. We are developing a modular simulation/visualization tool to effectively analyze and evaluate the effects these proposed paradigms will have on power system operations. Key research goals in- clude methods to assess transmission system capacity, pric- ing of transmission capacity, and development of criteria for an equitable and consistent comparison of alternative paradigms.

Parallel processing algorithms for dynamic response calculations of large power systems with detailed models are being developed. The work is based on algebraizing the differential-algebraic system of equations of the power system using the simultaneous-implicit method. Instead of parallelizing the LU method for solving the resulting system of linear equations at each time step using Newton's method, we use the conjugate gradient method. Use of preconditioners such as the ILU(s) speeds up the convergence. Further enhancement in speed-up is obtained by using the preconditioner only when the number of iterations increase. The general minimal residual (GMRES) method suitable for matrices that are not positive definite is used.

In this research, we compute trajectory sensitivities of the post-fault system with respect to prefault loading conditions and for a given set of contingencies. From this we compute whether the system is stressed and, if so, identify the critical machines. Results on a 17-machine IEEE test system are encouraging. In view of the fast computing power available these days, sensitivity theory offers an alternative to existing techniques for security assessment and preventive control.

In this research, we expand upon the MATLAB-based small-signal analysis formulation developed at the University of Illinois to include FACTS devices such as the static var compensator (SVC) and Thyristor Controlled Series Capacitor (TCSC). In particular, we will focus on controlling Hopf bifurcation through proper placement of these devices. Computation of the magnitude and frequency of the limit cycles using center manifold theory will be carried out.

We are using the asymptotic expansion theory for the ``inner'' and ``outer'' solutions of a singularly perturbed two time-scale system to systematically integrate the fast and slow subsystems in their respective time scales thus removing the ``stiffness'' of the original system. This is an alternative to using the integral manifold theory. The two approaches are compared in terms of their computational speed and convenience for simulation using the example of a synchronous machine subjected to a disturbance.

We are using interval matrix theory to see if the linearized model of a power system is Hurwitz stable with respect to variations of the elements of the matrix in a given interval. The initial application has been with respect to power system stabilizer (PSS) parameter variation, which can be expressed in a matrix polytope form. Multimachine ap- plication is now being done with loads being taken as perturbations.

Physically based preconditioners will be developed for fast nonlinear simulation of power systems using the general minimal residual (GMRES) iterative solver technique. It will be compared with the LU factorization method. Both will be developed on the MATLAB platform and integrated with the existing small-signal stability program and the transient energy function program. Ultimately, the idea is to develop a power system dynamics toolbox useful for R&D of small to medium sized systems.

The purpose of this project under the Indo-U.S. Science Cooperative Program is to collaborate in the area of small-signal analysis of large-scale power systems. Specifically, the topics to be addressed are the design of power system stabilizers, investigation of torsional oscillations, and computation of selected eigenvalues of the system. The ultimate goal is to bring out a research monograph in this area useful to the power engineering community. A preliminary set of lecture notes has been developed.

This project examines methods to eliminate sustained oscillations when they appear in power systems. An eigenvalue sensitivity approach is being tested to determine the effectiveness of operator controls that may eliminate oscillations. Effectiveness of discrete control actions, such as disabling a voltage regulator, is also being investigated.

This project is investigating the importance of data accuracy in the real-time control of power systems. Sensitivities of static and dynamic response calculations to input data are being examined and related to results such as security margins, transfer capability, stability limits, and economic dispatch.

Electrostatic spraying is a method of producing small particles and ions. These small particles and ions are emitted from liquids by electrostatic forces. Three different types of electrostatic spraying are the emission of ions from a conical surface, the emission of particles from a conical surface, and the formation of drops from an electrostatically formed jet. This project is designed to determine the conditions that allow each of these modes to exist. The conditions that cause the conical surface to become unstable and the conditions for the formation of a stable jet will be established. This will allow a calculation of the particle size produced by the electrostatic spraying to be calculated. A wide variety of liquids from perfect insulators to good conductors will be examined.

A solar electrical car is being designed and constructed by students to compete in a cross-country solar car race (SunRayce 1997) to be held in June 1997. Mechanical engineering considerations include the minimization of drag coefficient, rolling resistance, and weight. Electrical engineering considerations include optimizing the amount of power transferred from a solar array to storage batteries and maximizing the efficiency of the drive motor and the inverter that supplies its energy. All this must be done while producing an operating vehicle that conforms to the rules of the competition. This project involves approximately 100 students.

The goal of this program is to develop a software simulation for the 60-Hz magnetic fields in and surrounding a typical residence. Sources for these fields include household appliances, house wiring, and nearby power lines. The software will provide a dynamic tool for the realistic analysis of magnetic field distributions.

For improved braking and to detect problems, it is desirable that each car in a freight train have available a source of electrical energy. The goal of this project is to determine all the possible ways this energy can be supplied and to evaluate them. More detailed studies of the most promising schemes will be conducted.