Method and apparatus for measuring volt-amps reactive power using synthesized voltage phase shift

Abstract

In the present invention, true volt-amperes reactive power is measured by sampling instantaneous voltage values of voltage and substantially simultaneously instantaneous values of current of an arbitrary waveform, multiplying the voltage samples by the discrete Hilbert transform (corrected by a sign inversion) to obtain voltage samples of all frequencies shifted by -90.degree. over the bandwidth of the signal of interest and then multiplying this quadrature voltage sample by the corresponding current samples to obtain a true value for VARs, i.e., proportional to the instantaneous volt-amperes reactive power under all conditions. The sampling rate and the period of sampling determine the bandwidth and accuracy of the sampling technique. The invention can be implemented in software or in suitable hardware components capable of sampling and delaying the signal of interest.

Description

BACKGROUND OF THE INVENTION

This invention relates to the accurate measurement of voltage-ampere reactive power (VAR) and more particularly a method and apparatus for measuring VARs in a real world environment where the waveform of interest cannot be accurately approximated as a sinusoid.

Electrical utility system designs are based on the kilovolt-ampere (KVA) load to be served. The KVA load is regarded as the power expended under a given range of power factors. The power factor is defined as the ratio of current drawn by the load to the total current. The total current is defined in terms of kVARs. KVAR metering, or so-called phase-displaced metering, is an attempt to measure voltage and current in quadrature phase relationship to one another, under the assumption that the waveform is sinusoidal (i.e., has no harmonics). The VAR measurement derives from the unit of measurement of magnetizing volt-amperes and is an abbreviation for the phrase "volt-amperes reactive". It is conventionally defined as the voltage times the magnetizing current in amperes.

A theory of power measurement has been developed based on the use of VAR measurements, not all of which is technically correct. Moreover, traditional metering does not yield sufficiently-accurate VAR measurement to permit accurate system design and load management. Traditional measurement techniques involve measuring the current, measuring the voltage at the same time as the current is measured and then delaying the voltage measurements by a time equal to the phase difference of the fundamental frequency. As a consequence, a non-sinusoidal waveform is not accurately depicted and measured by conventional VAR measurement techniques, since a non-sinusoidal waveform contains harmonics of the fundamental which are not properly delayed. As a further consequence, traditional metering techniques may give wildly inaccurate VAR measurements, leading to inconsistent and misleading design and operation decisions, particularly where the voltage and current waveforms are subject to high distortion.

Traditional VAR calculation is described in the Handbook For Electricity Metering published by the Edison Electric Institute (8th Ed.) Chapter 9, pp. 173-196 (first Ed. 1912, latest Ed. 1981). Therein is described the traditional VAR metering techniques and theory of operation. It is of interest to note that the method of measurement of electrical quantities is dependent upon the rate, the cost and the required accuracy. What is needed is a metering technique which is highly accurate under all loading conditions.

SUMMARY OF THE INVENTION

According to the invention, true volt-amperes reactive power is measured by sampling instantaneous voltage values of voltage and simultaneously instantaneous values of current of an arbitrary waveform, multiplying the voltage samples by the discrete Hilbert transform (corrected by a sign inversion) to obtain voltage samples of all frequencies shifted by -90.degree. over the bandwidth of the signal of interest and then multiplying this quadrature voltage sample by the corresponding current samples to obtain a true value for VARs, i.e., proportional to the instantaneous volt-amperes reactive power under all conditions. The sampling rate and the period of sampling determine the bandwidth and accuracy of the sampling technique. The invention can be implemented in software or in suitable hardware components capable of sampling and delaying the signal of interest.

The invention will be better understood by reference to the following detailed description in connection with the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a waveform diagram showing voltage as a function of time and current as a function of time for an ideal sinusoidal signal.

FIG. 2 is a block diagram of a meter according to the invention.

FIG. 3 is a schematic block diagram of an implementation of a meter in accordance with the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Referring to FIG. 1, there is shown a pair of waveforms 10, 12 representing voltage v(t) and current i(t) of an electric power system. In the depiction, the voltage and current are nearly sinusoidal and exhibit no phase difference therebetween. In reality, the waveforms are likely to be distorted, and there may be a phase difference between voltage and current. In accordance with the invention, the voltage and the current are each sampled substantially simultaneously. For example, there is sampling of each waveform at sample times t=0 through t=31 over a time of two periods of the fundamental frequency of the waveforms.

Referring to FIG. 2, there is shown a block diagram of a meter 14 according to the invention. The meter 14 includes a voltmeter 16 across a load of interest L and an ammeter 18 in series with the load L (or coupled in shunt across a known low resistance R in series with the load L) and adapted to measure instanteous voltage and instantaneous current substantially continuously. The meter continuous outputs v(t) and i(t) are applied to sampling circuits 20 and 22, respectively, represented by switches 24, 26 and hold elements 28, 30. The switches are operated synchronously by a clock 32 so that the resultant samples v[n] and i[n] are synchronous and in phase.

The samples v[n] are applied to an all frequency phase shifter 34 according to the invention, as hereinafter explained, and the samples i[n] are time delayed as represented by a sample delay 36 to remain synchronous with the voltage samples v[n].

The output of the all frequency phase shifter 34 is the voltage samples v[n] delayed by 90 degrees, or v.sub.-90 [n]. The samples i[n] and v.sub.-90 [n] are multiplied together in a multiplier 38, the product of which is scaled through an amplifier 40 by an appropriate gain factor K.sub.1, the output of which is the instantaneous VARs 42. The VARs samples are summed over the sampling period N and scaled to obtain the average VARs 44 for the sampling period. KVAR-HOURS 46 is extracted from the instantaneous VARs 42 samples or from the average VARs by appropriate scaling, taking into account the frequency of sampling, the number of samples, the duration of the measurement period and by simply converting VARs to KVARs.

According to the invention, the all frequency phase shifter is a discrete Hilbert transform processor wherein the output is multiplied by negative 1. The coefficients for a simple multiplier realization of such a transform processor are given in Oppenheim and Schafer, Digital Signal Processing. (Prentice-Hall, 1975), page 360, FIG. 7.10. The coefficients are taken from the normalized impulse response of an ideal Hilbert transformer over an appropriate period corresponding to the sampling period before and after the instant of interest. In a specific embodiment, the period for sampling is one complete cycle plus one-half cycle before the cycle of interest and on-half cycle after the cycle of interest. The coefficients are the core of a ninety-degree phase shifter. Only odd values of samples are employed with the ninety-degree phase shifter, since the even values correspond to coefficients of value zero. As with any sampled-data system, aliasing of terms is a concern. Accordingly, the bandwidth of interest is selected to eliminate concerns about values aliased into the passband of interest.

The ninety-degree phase shifter is a realization of the following equation:

v.sub.-90 [n]=3v[n-5]+5v[n-3]+15v[n-1]-15v[n+1]-5v[n+3]-3v[n+5](1)

The foregoing equation is depicted in FIG. 3 as element 34 which comprises negative 1 multiplier 50, unit delay 52, dual unit delays 54, 56, 58, 60, 62, coefficient multipliers 64, 66, 68, 70, 72, 74, and parallel adders 80, 82, 84, 86, and 88. The all frequency delay 36 is a six unit delay such that multiplier 38 combines the multiplicand and multiplier at the midpoint of the delayed time interval of the ninety degree phase shifter 34. The gain factor of the amplifier K.sub.1 is 1/15 to compensate for the integer gain of the coefficient multipliers of element 34. The resultant output at 42 is an accurate representation (within sampling error) of instantaneous VARs delayed by six time units or six sample times. This result compares favorably with any other method of measuring VARs, and it exhibits the same level of accuracy for any waveform exhibiting distortion and having no substantial harmonics above one-half the sampling rate. Although the Hilbert transform is infinite, a representation containing six non-zero coefficients (or equivalently 13 zero and non-zero coefficients) provides adequate precision over the bandwidth of interest.

Other related values are readily derived from the instantaneous VARs measurements. For example, the average VARs for the measurement period, i.e., two seconds or N=128 samples is given by: ##EQU1## A suitable adder element 43 (FIG. 2 or FIG. 3) performs this function. The VAR-HOURS is computed from VARS(avg) in accordance with the following relationship:

VAR-HOURS=VARS(avg) (N/f.sub.s) (1/3600) (3) where f.sub.s is the sampling frequency.

Elements 45 and 47 provide the appropriate scaling.

Appendix I, attached hereto and made a part hereof, is a listing of a software implementation of the method according to the invention as applied to a meter having an internal microprocessor capable of performing calculations on values representing sample voltage values and sample current values. This software routine is driven by the interrupts in the system used to acquire samples. The routine performs a portion of its calculation after each sampling in order to distribute the calculation burden of the processor. The result is a calculation which is based on a ninety-degree phase shift for all frequencies of interest. Also included in the routine are calculations of power and other quantities. In calculating power, a random delay is included in the data stream. This is equivalent to randomizing the data samples in order to assure that the steady-state waveform is accurately sampled.

While the invention has been particularly shown and described with reference to preferred embodiments thereof, it will be understood by those skilled in the art that the foregoing and other changes in the form and details may be made therein without departing from the spirit or scope of the invention. ##SPC1##

Claims

1. A method for measuring reactive power in an electrical load in an electrical system, said method comprising the steps of:

electromagnetically coupling a voltage probe across the electrical load;

electromagnetically coupling a current probe to the electrical load;

measuring a time-varying voltage across the electrical load using said voltage probe, said voltage probe outputting a voltage signal representing said time-varying voltage;

measuring a time-varying current flowing through the electrical load using said current probe, said current probe outputting a current signal representing said time-varying current;

sampling substantially simultaneously and instantaneously said voltage signal and said current signal at a sampling frequency to produce a plurality of voltage samples representative of said voltage signal and a plurality of current samples representative of said current signal;

shifting the phase of frequency components of said plurality of voltage samples by -90.degree. C. relative to said plurality of current samples for substantially all frequency components having a frequency less than one-half of said sampling frequency to produce a plurality of quadrature voltage samples; and

multiplying each sample of said plurality of quadrature voltage samples by each corresponding sample of said plurality of current samples to produce a plurality of reactive power samples, each of which is a measure of instantaneous reactive power in the electrical load.

2. A method according to claim 1, further comprising the steps of:

averaging said plurality of reactive power samples to obtain a value proportional to average reactive power in the electrical load; and

providing said value proportional to average reactive power to an output utilization device.

3. A method according to claim 2, wherein said sampling step comprising sampling over a period of at least one cycle of a fundamental frequency of interest and at least one quarter cycle prior to and at least one quarter cycle following a period of interest.

4. A method according to claim 1, wherein said sampling step comprises sampling over a period of at least one cycle of a fundamental frequency of interest.

5. A method according to claim 1, further comprising the steps of:

normalizing at least one of said plurality of reactive power samples to obtain a calibrated value for instantaneous reactive power in the electrical load; and

providing said calibrated value for instantaneous reactive power to an output utilization device.

6. A method according to claim 1, wherein the step of sampling substantially simultaneously instantaneous voltage and instantaneous current is done over a period of at least one cycle of a fundamental frequency of interest and at least one quarter cycle prior to and at least one quarter cycle following a period of interest.

7. A method according to claim 1, wherein at least one said voltage signal and said current signal is not purely sinusoidal in amplitude.

8. A method according to claim 1, further comprising the step of providing said measures of instantaneous reactive power to an output utilization device.

9. A meter for measuring reactive power in an electrical load in an electrical system, comprising:

a voltage probe electromagnetically coupled across the electrical load, for measuring a time-varying voltage across the electrical load, wherein said voltage probe outputs a voltage signal representing said time-varying voltage;

a current probe electromagnetically coupled to the electrical load, for measuring a time-varying current flowing through the electrical load, wherein said current probe outputs a current signal representing said time-varying current;

sampling means coupled to said voltage probe and to said current probe, for sampling substantially simultaneously and instantaneously said voltage signal and said current signal at a sampling frequency to produce a plurality of voltage samples and a plurality of current samples;

shifting means coupled to said sampling means, for shifting the phase of frequency components of said plurality of voltage samples provided to said shifting means by said sampling means by -90.degree. C. relative to said plurality of current samples for substantially all frequency components having a frequency less than one-half said sampling frequency, to produce a plurality of quadrature voltage samples; and

combining means coupled to said sampling means and said shifting means, for multiplying each of said plurality of quadrature voltage samples provided to said combining means by said shifting means by each corresponding sample of said plurality of current samples provided to said combining means by said sampling means to produce a plurality of reactive power samples, each of which is a measure of instantaneous reactive power in the electrical load.

10. A meter according to claim 9, further comprising:

averaging means coupled to said combining means for averaging said plurality of reactive power samples provided to said averaging means by said combining means to obtain a value proportional to average reactive power in the electrical load;

scaling means coupled to receive said value from said averaging means, said scaling means for multiplying said value by an inverse of said sampling frequency to produce a VAR-hours value representing the number of VAR-hours of reactive power in the electrical load; and

an output utilization device coupled to said scaling means, wherein said scaling means outputs said number of VAR-hours to said output utilization device.

11. A meter according to claim 10, wherein said output utilization device records said number of VAR-hours for utility service billing purposes, and wherein said power meter is located proximate the electrical load.

12. A meter according to claim 9, further comprising:

averaging means coupled to said combining means for averaging said plurality of reactive power samples provided to said averaging means by said combining means to obtain a value proportional to average reactive power in the electrical load; and

an output utilization device coupled to said averaging means, wherein said averaging means outputs said value proportional to average reactive power to said output utilization device.

13. A meter according to claim 9, wherein said shifting means is a digital filter with a transfer function which is a digital approximation of a Hilbert transform.

14. A meter according to claim 9, wherein at least one of said voltage signal and said current signal is not purely sinusoidal in amplitude.

15. A meter according to claim 9, further comprising an output utilization device coupled to said combining means, wherein said combining means outputs said measures of instantaneous reactive power to said output utilization device.

16. A meter according to claim 9, wherein said meter is implemented using a digital signal processor.

17. A meter for measuring reactive power in an electrical load in an electrical system, comprising:

a voltage probe electromagnetically coupled across the electrical load, for measuring a time-varying voltage across the electrical load, wherein said voltage probe outputs a voltage signal representing said time-varying voltage;

a current probe electromagnetically coupled to the electrical load, for measuring a time-varying current flowing through the electrical load, wherein said current probe outputs a current signal representing said time-varying current;

a sampling circuit coupled to said voltage probe and said current probe, said sampling circuit sampling substantially simultaneously and instantaneously said voltage signal and said current signal at a sampling frequency to produce a plurality of voltage samples and a plurality of current samples;

a shifter coupled to said sampling means, said shifter shifting the phase of frequency components of said plurality of voltage samples provided to said shifter by said sampling means by -90.degree. relative to said plurality of current samples for substantially all frequency components having a frequency less than one-half said sampling frequency, to produce a plurality of quadrature voltage samples; and

a multiplier coupled to said sampling circuit and said shifter, said multiplier multiplying each of said plurality of quadrature voltage samples provided to said multiplier by said shifter by each corresponding sample of said plurality of current samples provided to said multiplier by said sampling circuit to produce a plurality of reactive power samples, each of which is a measure of instantaneous reactive power in the electrical load.

18. A meter according to claim 17, further comprising:

averaging means coupled to said multiplier for averaging said plurality of reactive power samples provided to said averaging means by said multiplier to obtain a value proportional to average reactive power in the electrical load; and

an output utilization device coupled to said averaging means, wherein said averaging means outputs said value proportional to average reactive power to said output utilization device.

19. A meter according to claim 17, further comprising:

averaging means coupled to said multiplier for averaging said plurality of reactive power samples provided to said averaging means by said multiplier to obtain a value proportional to average reactive power in the electrical load;

scaling means coupled to receive said value from said averaging means, said scaling means for multiplying said value by an inverse of said sampling frequency to produce a VAR-hours value representing the number of VAR-hours of reactive power in the electrical load; and

an output utilization device coupled to said scaling means, wherein said scaling means outputs said number of VAR-hours to said output utilization device.

20. A meter according to claim 17, wherein said shifter is a digital filter with a transfer function which is a digital approximation of a Hilbert transform.

21. A meter according to claim 17, wherein at least one of said voltage signal and said current signal is not purely sinusoidal in amplitude.

22. A meter according to claim 17, further comprising an output utilization device coupled to said multiplier, wherein said multiplier outputs measures of instantaneous reactive power to said output utilization device.