Gerald L. Park

Modification of Power System Operation for Significant Wind Generation Penetration

The purpose of this paper is to discuss modification of unit commitment, economic dispatch, regulation and frequency regulation controls when the level of wind generation capacity is significant. A wind farm penetration constraint is determined, that limits worst case wind generation change from an array due to a thunderstorm to be less than the worst first contingency loss of conventional generation resource or commitment. The farm penetration constraint only acts as an indicator that additional spinning reserve, load following, and unloadable generation capability is required through adjustment of unit commitment and AGC controls if the farm penetration constraint was violated. A discussion of the methodology, costs, and benefits of changing unit commitment when WECS generation is significant (and either satisfies or violates this farm penetration constraint) is then discussed. A further discussion of the modification of regulation and economic dispatch controls to exploit the changes in response rate capability provided by the unit commitment is also discussed. A modified echelon penetration constraint that limits instantaneous rate of change and change from a wind array that must be handled by governor frequency regulation and regulation controls. This constraint is imposed to limit cycling of units which can incur additional operating and maintainance costs on conventional steam units and possibly safety concerns on nuclear units.

A Modified Unit Commitment and Generation Control for Utilities with Large Wind Generation Penetrations

A modified unit commitment is proposed in this paper that would be updated on three different cycles (daily, quarter-hourly and every minute). The 24 hour updated unit commitment would include the effects of load and slow trend wind power change that could be predicted 24 hours ahead. The quarter- hourly updated unit commitment would handle the fast trend and cyclic change in wind power that could be predicted one hour ahead and would commit peaking, regulation, and quick pickup and possibly economic units to handle these wind power changes.

Simulation and Assessment of Wind Array Power Variations Based on Simultaneous Wind Speed Measurements

A wind model for relating wind speeds based on simultaneous wind speed measurements at several sites is developed for both meteorological events and turbulence wind conditions. The wind model is used to perform an analysis and simulation of wind power variations from single and multiple wind turbine arrays for meteorological events and turbulence. The results indicate that serious reduction in operating reliability and economy can occur for meteorological event induced wind power variations when total wind power generation capacity exceeds spinning reserve levels on a utility. The effects of turbulence induced variations are shown to be small for 5% wind penetration on a large utility but quite serious for the same penetration level on a small utility.

Voltage reduction as a means of reducing distribution load

Most electric utilities have emergency procedures for reducing system load when their generation and outside purchases cannot meet the expected load and reserve requirement. One such procedure reduces voltage by means of transformer tap changes or voltage regulators at selected substations and circuits. This paper is not concerned with the means by which this is done but rather with the selection of circuits for voltage reduction. For some time, the rule of thumb in the industry has been that a 1% voltage change produces 1% total power change. At the present time there is considerable questioning of how much load reduction is actually gained by voltage reduction. In an attempt to clarify the behavior of different types of substations, a series of tests was performed at three different types of substations; residential, commercial and a combination of residential and commercial. These tests were performed in the afternoon and evening in the winter, spring and summer utilizing fast instrumentation developed by Karl Andrews and G. L. Park.2

Methods of Reducing Wind Power Changes from Large Wind Turbine Arrays

Previous results have [1,2,3] been concerned with establishing (1) whether operating problems could exist when WECS generation is significant and (2) the proper modification of unit commitment, regulation, and economic dispatch required to provide sufficient system security and alleviate the operating problems caused by WECS generation changes. This paper discusses methods of reducing the WECS generation change through selection of the wind turbine model for each site, selection of an appropriate siting configuration, and wind array controls. An analysis of wind generation change from an echelon and a farm for passage of a thunderstorm is presented to establish the factors concerning the wind turbine model and ; siting configuration that contribute to these variations. Detailed simulation results indicate more. precisely how these factors can be exploited to. minimize the WECS generation changes observed. Reduction of the wind generation change over ten minutes is shown to reduce the increase in spinning reserve, unloadable generation and load following requirements on unit commitment when significant WECS generation is present and the farm penetration constraint is satisfied. Controls on the blade pitch angle of all wind turbines in an array or a battery control are shown to reduce both the wind generation change out of an array and the effective farm penetration in anticipation of a storm so that the farm penetration constraint may be satisfied.

Model identification using tie line power and frequency measurements

The identification of dynamic models which relate power and frequency deviations on a tie line of a power system is investigated. The identification problem is posed and three identification algorithms are presented which produce least squares models with different structural properties. Model order is determined by applying residual and system structure tests to a sequence of models of increasing order. These tests indicate the model order for both equivalent realizations and predictive models. Equivalent realizations are identified on one data set and then their performance as a dynamic equivalent is evaluated on a second data set. These equivalent realizations are also used to predict frequency in an iterative frequency prediction algorithm. Predictive models are also identified and their performance as frequency predictors is evaluated using a direct prediction algorithm. The identification of dynamic equivalents provides information about the structural properties of power systems. The use of dynamic equivalents and predictive models for frequency prediction indicates the tradeoff in accuracy vs the prediction interval which can be obtained using these least squares algorithms and the measurement device presently available.

Analysis and Simulation of Storm Induced WECS Generation Changes on System Operation

The effects of large rapid changes in generation from large arrays of wind turbine generators on the operation of automatic generation control (AGC) and frequency regulation will be assessed. The theoretical worst case change and rates of change of generation from an array of wind turbines due to passage of a thunderstorm front is given for a coastal farm. Constraints on the penetration of the portion of an echelon and farm that are affected by the front are derived so that power variation rates from a farm satisfying these constraints will not exceed the response rate capability of a typical system. These penetration constraints would eliminate the occurrence of excessive frequency excursions and violations of NAPSIC performance requirements on automatic generation control. The penetration constraints effect on the size of an array and the maximum power variation rates allowed on a particular system are discussed. The effects of simultaneous load and Wind Electric Conversion System (WECS) generation changes on automatic generation control, frequency regulation, and economic dispatch are then determined via simulation. It is shown that the total of load and WECS generation change in a ten minute interval must be less than load following capability for that period or a severe saturation problem occurs. A cycling problem is shown to occur if storm induced WECS generation change will cause frequency deviations that exceed governor deadband.

State Estimation of a Simulated Turbine Generator

A model of a single steam turbine generator is developed which includes the effects of random load fluctuations observed in interconnected power systems. This model is used in the design of an on-line state estimator for an individual generator. A digital simulation of the generator model and state estimator indicates the state estimation procedure is feasible and can be implemented on a small control computer.

A study of frequency prediction for power systems

A frequency predictor is identified from simulated tie-line measurements of frequency and power. The performance of the predictor on the simulation and on actual measurements is shown to be identical out to a one second prediction interval. Frequency prediction beyond one second is shown to be poor because the predictor no longer identifies the dynamic relationship between frequency and power and because good frequency prediction at greater than one second ahead of real time would require accurate prediction of random load changes. The predictor is not sensitive to changes in sampling rate and to measurement noise levels as long as the measurement noise is very small with respect to the excursions being measured. Finally, the predictor is tested on different size and types of disturbances. The error in prediction is shown to be proportional to the size of the disturbance. The predictor has large errors for several seconds after the loss of a generator or major tie-line. The predictor tracks the new system dynamics and resumes accurate frequency prediction approximately five seconds after the disturbance occurred.

Derivation of a suboptimal linear filter

A linear filter using sample extrapolation and averaging is derived and tested. Extensions to include correlated observation noise and incomplete measurement are discussed. The filter is easily implemented for low-order linear plants but is inferior to the Kalman filter for all but the simplest cases.