A.D. Patton

Real power transfer capability calculations using multi-layer feed-forward neural networks

This paper proposes a neural network solution methodology for the problem of real power transfer capability calculations. Based on the optimal power flow formulation of the problem, the inputs, for the neural network are generator status, line status and load status and the output is the transfer capability. The Quickprop algorithm is used in the paper to train the neural network. A case study of the IEEE 30-bus system is presented demonstrating the feasibility of this approach. The new method will be useful for reliability assessment in the new utility environment.

Optimal Thermal Generating Unit Commitment

The paper describes a new method of scheduling thermal generating units to achieve minimum operating costs including both running and start-up costs while at the same time maintaining a desired level of system security.

Operating Considerations in Generation Reliability Modeling-An Analytical Approach

The paper presents a new analytical approach to the calculation of generating system reliability indices. The new approach makes it possible to relax idealizing assumptions and to explicitly model the effects of operating considerations such as: (1) unit duty cycles reflecting load cycle shape, reliability performance of other units, unit commitment policy, and operating reserve policy; (2) start-up failures; (3) start-up times; and (4) outage postponability. The models presented can also be used to consider the effects of basic energy limitations and to give production cost estimates.

Protection System Reliability Modeling: Unreadiness Probability and Mean Duration of Undetected Faults

A reliability model of a system and its associated protection system is described. The protection system detects the presence of faults and isolates the faulted equipment so as to prevent damage and minimize the effect of the faulted equipment on the operation of the rest of the system. If the protection system does not respond and the faulted component is not disconnected, a backup protection system comes into operation. The backup operation isolates a larger segment and disturbs the system more than if the designated protection system had responded appropriately. An example is the protective relay systems for electric power transmission and distribution systems. A measure of system unreadiness is defined and suitable expressions are derived for this index and the mean duration of undetected faults. The unreadiness probability is sensitive not only to the failure rate of the protection system but also to the failure rate of the system, the pdf of interval between inspections and its mean value. The unreadiness probability estimated from a given system cannot be used for another system having a similar protection system. The protection system failure rate can, however, be used to obtain the unreadiness probability.

Models and concepts for power system reliability evaluation including protection-system failures

Abstract Models, concepts and methods for incorporating the effect of protection-system failures into the reliability evaluation of power transmission networks are presented.

Power system reliability evaluation using learning vector quantization and Monte Carlo simulation

Abstract Artificial Neural Networks (ANN) based on the Learning Vector Quantization (LVQ) algorithm have received considerable attention as pattern classifiers. This paper proposes a new method for power system reliability evaluation combining Monte Carlo simulation and LVQ which greatly reduces the computing burden of the loss of load probability calculation compared to Monte Carlo simulation only. A case study of the IEEE RTS system is presented demonstrating the efficiency of this approach.

Short-Term Reliability Calculation

The paper suggests a reliability criterion for use in adaptive control of power systems. A generalized method of forecasting the reliability or security of an operating system on a short- term basis is given. Also given are specific probabilistic models for use in forecasting the sufficiency of spinning generation capacity.

Determination and Analysis of Data for Reliability Studies

Methods for transmission and distribution system reliability analysis currently being developed and used require that the expected values of component outage rates and outage durations be known. The paper describes how the required component parameters can be estimated from field data. Also described are methods of placing confidence limits on the parameters estimated. The methods presented were developed through a study of six years of field records provided by a utility company. Sample expected outage rates and outage durations for various components such as transmission lines, substation transformers, and circuit breakers are presented. Also presented are data which tend to confirm the distributional assumptions made in system reliability analysis methods.

Modeling and evaluation of the system reliability effects of direct load control

Direct load control, a form of load management in which portions of the system load are under the direct operational control of the utility, is discussed. The load can be modified, within limits, to match the available generating capacity, thereby minimizing the occurrence of uncontrolled load loss. Accordingly, direct load control renders the system load a function of available system capacity. Therefore, accurate modeling of system reliability effects of direct load control requires that the temporal correlation between system load and available generating capacity be preserved. This is most readily done using Monte Carlo simulation. A Monte Carlo simulation model and results obtained with it are described. >

A production costing methodology for evaluation of direct load control

Direct load control (DLC) is a form of load management in which portions of the system load are under the direct operational control of the utility. Thus, the load can be modified, within limits, to match the available generating capacity, thereby minimizing events of uncontrolled load loss. It is shown that operating cost reductions produced by exogeneous models of DLC are different from those produced by dynamic models of DLC. The dynamics of DLC and an equal incremental cost scheme are described and then coupled with a Monte Carlo simulation model of the operation of thermal-based electric utilities. Various production cost measures produced by exogeneous and dynamic models of DLC are presented and discussed. >