GB2143955A - Electronic apparatus for metering and measuring electrical energy consumption in a manner which simulates the same measurement as effected by a rotating disc electromechanical meter - Google Patents

Abstract

Apparatus whose measurement means consists solely of electronic components, measures the power taken by an electrical load connected to an alternating source of electricity supply in such a manner that the power sensed or energy accumulated is in agreement with the same measurements as performed by a conventional rotating disc electromechanical meter. Fundamental frequency filtering means 3, 6 when applied to either one or both supply voltage and load current ensure that the electronic circuitry performs the appropriate measurements in determining power consumption. A single sample per mains cycle applied to load current and supply voltage is sufficient to measure power consumption, energy consumption accumulation, and produce accumulated energy pulses appropriate to measurement of real power, true power and volt-ampere power consumed by the load. <IMAGE>

Description

SPECIFICATION Electronic apparatus for metering and measuring electrical energy consumption in a manner which simulates the same measurement as effected by a rotating disc electromechanical meter The present invention relates to the measurement of electrical energy by electronic circuitry. The traditional means of measuring the amount of electrical energy consumed by a load is to use the Ferraris Wheel electromechanical meter. This kind of meter is used predominantly for the measurement of electricity consumption by electricity supply authorities in order to charge their customers for electricity consumed.

The Ferraris Wheel meter is an electromechanical device which drives a gear chain from a rotating disc which turns in sympathy with the power utilisation. The gear chain turns indicator discs which allow a totalised reading of Kilowatt hours to be taken at desired intervals. The difference between consecutive readings indicates the total energy consumed in the period.

The Ferraris Wheel meter is a robust highly developed and effective means of measuring electrical energy. As well as being used for electricity supply metering it is often employed for the measurement of totalised energy in any situation which demands an assessment of energy utilisation. Examples are the measurement of electrical machine consumption in factories, calculation of factory process costs where electricity is used for heating, refrigeration etc., energy efficiency measurements in experimental and teaching situations measurements of electrical energy delivery from generators and consumption by motors.

The Ferraris Wheel meter has good accuracy over a wide range of power consumption and it is insensitive to mains power interruption. Where remote monitoring of power utilisation is required pulsing Kilowatt hour meters are available which are arranged to produce relay contacts or some other electrical signal every time a preset unit of energy is consumed, such as 1.0 Kilowatt hour.

However, advances in the state of electronics have made it possible to monitor electricity consumption by totally electronic means. The advantage of solid state circuitry is that the sensing of electricity flow can be more easily integrated into systems which require remote communication or interrogation of the metering device. Also as the state of the art of electronics advances the all electronic meter will become competitive in performance in price, size and convenience with the Ferraris Wheel meter.

There has been much work conducted to design suitable circuitry for all-electronic measurement of electrical power, as well as the integration of that electrical power to produce a totalised energy measurement. A simple and effective means of achieving this is well described in UK Patent GB 2 041 588 B. The single sample approach described gives accurate results on linear loads but is inaccurate for non-linear load currents.

In order to clearly show the reasoning behind this invention it is desireable to review some fundamental principles of electrical power and its measurement.

To assess power in a circuit it is necessary to produce a measureble physical quantity proportional to the power. Various means have been used to perform this function. Since true instantaneous power, p, is proportional to the product of instantaneous current, i, and instantaneous voltage, v, the power in a load at any instant is given by the expression p = vi Average power P is the effective power averaged over one or several cycles of the alternating voltage supply, which discounts short-term variation of instantaneous power through the mains voltage cycle.

We compute average power over one mains cycle as the following expression.

vidt where T is the period of the supply voltage, equal to #s 2" çs being the supply frequency measured in radians per second.

In electronic terms the above equation indicates that we should have a device sensitive to voltage across the circuit, another sensitive to current in the circuit, produce an electronic multiplication of the two quantities and then average the resultant quantity which is varying in sympathy with instantaneous power with a time constant equal to the duration of several cycles of the supply voltage. Alternatively it is possible to average the output of a single device sensitive to instantaneous circuit power.

Such an arrangement is shown in Fig. 1. In this arrangement instantaneous power is computed by an analogue multiplier sensitive to voltage and current in the load. The output of this circuit is averaged to provide a quantity proportional to true average power in the load. The precise means of performing the multiplication and averaging functions is immaterial and consists of any analogue, digital, or combined means of performing these functions. It is possible to use operational amplifiers, pulse frequency modulation, digital circuitry or a combination of these methods. Sensing of the voltage and current in the load can be performed by a variety of means such as potentiometers, current transformers, Hall Effect voltage, current, or power sensors or any other means.However the average power may be derived, the function of the sensors and circuitry can be reduced to the representation of Fig. 1.

It is of interest to present the mathematical function realised by the representation of Fig. 1.

Although the supply voltage is nominally sinusoidal let us represent the supply voltage as a function v.

= = v(t) and the load current as a general current i, is = i(t) We can expand v8 and is as a fourier expansion of harmonics, vn, of the basic frequency , of the mains voltage.

So, omitting for simplicity phase angles, Vs = v0 + v1 cos a;,t + + v, cos##st...

Also is = i0 + i1 cos at + + im cosm w,t..

The instantaneous power is computed by the multiplier or power sensing component and averaged by the averaging circuit. Since supply frequency co, is sensibly constant we can drop the subscript, s, and for simplicity represent supply frequency by constant co.

Instantaneous power p(t) is thus

And average power, P, is

The general term after performing the double summation function will be

And Pnm will only have non zero value if n and m are equal or either is equal to zero. Hence, the evaluation of the power function is

This is the function evaluated by the true power meter, including any dc terms present arising from the v0 and i0 components in the preceeding power formula.

However, it is the object of this invention to compute power as charged by an electricity supply authority. This is almost always computed by the Ferraris Wheel meter and we now examine the function performed by this meter.

The Ferraris meter computes power by the interaction of two fluxes which are instantaneously proportional to both current and voltage in the load. A flux is generated by the load current passing through a current coil and a flux at 90 is generated by a coil connected across the supply voltage. The interaction of these two fluxes through the medium of induced eddy currents in a conducting (aluminium) disc cause a torque proportional to the product of these two fluxes. Rotation of the wheel, (and hence meter reading) is proportional to the integrated or averaged product of these two fluxes over time, and consequently the energy consumed.

For the Ferraris Wheel meter we have again

However, the mechanism of the Wheel can not induce torque dependent on dc compononts of either current or voltage, hence only the following terms can be sensed.

Consequently, the Ferraris Wheel responds only to the fundamental component at mains frequency plus similar interacting harmonics of the mains voltage and load current. DC terms are ignored (and no cross product terms are sensed). This means that a modification of the circuit of Fig. 1 having AC coupling of current and voltage sensing into the multiplier would simulate the Ferraris Wheel. A low pass filter having a similar response to the Ferraris Wheel meter is required at the Multiplier output as shown in Fig. 2.

But if we examine the above expression for power and assume that the mains voltage is sinusoidal it is possible to simplify the expression for average power even further so that

We can ensure that only the fundamental of mains voltage is present by introducing a low pass filter with a sharp cut off above the mains frequency Ç on the voltage sense input. We now have a circuit as represented by Fig. 3. The output of the multiplier after averaging reflects only the mains frequency voltage fundamental multiplied by the fundamental of the load current component. It is possible therefore without changing the systems output in any way to insert a fundamental frequency filter also on the current sensing voltage as in Fig. 4. The final output of this circuit (after averaging) remains the same as that of Fig. 3. The circuit of Fig. 4 differs from Fig. 3 in that at the multiplier output no power components above 2z can be present since fundamental filtering is present on both sensing inputs.

Now the output of the multiplier in Fig. 4 is given by the expression p(t) = v cosot X i, coswt and the average fundamental power at the system output is given by

When the voltage and current are out of phase by an angle f the well known expression for ac power in a linear load results P=VI cosf Where V and I are the RMS values of the inputs to the multiplier, equivalent to the fundamental components of supply voltage and load current.

So by sampling voltages varying in proportion to the mains voltage fundamental and the load current fundamental both at their peaks, and also sensing the phase between these fundamentals, the average power over one or several cycles can be computed, as illustrated in Fig. 5. The power that is sensed by this means will be equal to the power sensed by the Ferraris Wheel meter apart from a small error due to harmonic distortion of the mains nominally-sinusoidal voltage. This will be true for all types of current load, both sinusoidal and non-sinusoidal, that is for both linear and non-linear loads.

Since this Ferraris Wheel simulation employs simply sensing and sampling of two sine wave voltages we can measure the voltage of the two sensed voltages at thier peak as in Fig. 5, or the voltage of both at the peak of one, preferably the voltage wave-form peak since the electricity supply voltage is always close to a sinusoid. This latter approach is illustrated by Fig.

6.

Due to the fact that the point of sample of the current waveform is shifted by + from an inphase position and the input waveform is sinusoidal, a sampling of i, at the voltage peak produces I cos f directly without further computation. The multiplier in this circuit needs only to respond to two sampled inputs to produce the desired fundamental power.

Features and advantages of the present invention will become apparent from the following more detailed description of a particular embodiment thereof, given by way of example with reference to the accompanying drawings in which, Figs. 1-6 are diagrams illustrating the principles leading to the embodiment, as already described.

Fig. 7 is a circuit diagram of a circuit to simulate the implementation of the Ferraris Wheel meter as used in a complete embodiment of a pulsing electronic Kilowatt hour meter.

In the embodiment of Fig. 7 the current is sensed by a current transformer 1.

The voltage produced across the load resistor 2, is proportional to the instantaneous ac current flowing in the load. This voltage is filtered by an active filter, 3, to leave only the 50Hz fundamental frequency as output of the filter.

This active filter has a gain of approximately 5 and the output is a sine wave varying in sympathy with the fundamental component of the load current.

The output of the filter is ac coupled to an amplification stage. The ac coupling removes any active filter dc offset. The amplifier stage, 4, has switched variable gain to allow for auto-ranging of gain over 3 decades, to measure current varying typically from 1 to 1000 amps load current.

The overall gain of the two stages in series is conveniently set to unity, X 10 and X 100 by appropriate resistors in the attenuation amplifier, 4. An appropriate output swing of the amplifier is 5 volts peak to peak. In this particular implementation CMOS amplifiers 7611 are employed.

The filter employed has an intrinsic phase shift of 90' at 50 Hz but since the mains voltage remains very close to 50 Hz at all times the phase shift remains sensibly constant and predetermined.

The phase reference provided by the ac voltage is detected in this implementation by the circuit, 5. This is directly connected to the mains to produce a negative output pulse with its leading edge occuring at the zero crossing of the mains. Due to the 90 phase shift on the current sensing voltage this point corresponds, conveniently to the peak of the current sense sinusoid for resistive loads.

As a consequence where there is a phase shift of the current the magnitude of the current sinewave sensing voltage is proportional to I cos + at the time the voltage zero crossing occurs.

The circuit, 6, detects a voltage which is proportional to the mains voltage by process of half wave rectification and dc low-pass filtering. This produces for convenience a dc voltage of 5 volts for a mains voltage of 255 volts ac. This is the input, V, required to complete the multiplication VI cos +, and is very close in practice to the value that would be determined from the fundamental component of the mains voltage, since mains voltage is close to sinusoidal.

In this particular embodiment digital samples of I cos +, and V are derived from the two input analogue channels of the processor, 7, in this case an Intel 8022 microprocessor. In this embodiment the two analogue voltages are combined by a process of analogue to digital conversion of I cos + and V with appropriate sampling time being dictated by the TO voltage zero crossing input. The 8022 processor multiplies the two digital samples to derive energy within the + cycle and accummulates each contribution, the resultant output being a pulse for every increment (say 1.0 KWh) of energy consumed as well as producing another digital signal indicative of the average power.

Another of many possible alternative embodiments would be to use a separate analogue to digital conversion device electronically switching between I cos 9 and V signals, and providing digital sampled values to an 8035, or 8748 or any other similar digital processing circuitry.

In the particular embodiment described we show how the processor, 7, can control the analogue switch, 8, to auto-range the measurement circuitry for improved accuracy over several decades of power variation via the 4066 analogue switch, 8.

An important feature of the amplifier circuit is that it has a linear gain characteristic. This means that analogue to digital sampling errors will be linearly distributed when measuring a varying load and will consequently average out over a period of time. A logarithmic amplifier characteristic does not have this feature and consequently is avoided in favour of a linear amplifier, with auto-ranging gain control provided by the processor. It should also be clear that the analogue to digital sampling device must also have a linear characteristic to preserve the averaging feature.

The essential feature of this particular embodiment is that it demonstrates one particular embodiment of our invention. This embodiment simulates electronically, by the use of a fundamental frequency filter, (in this case) on the load current only, together with a sensing of voltage phase and magnitude the amount of electricity consumption shown by the Ferraris Wheel meter. The embodiment shown will effect this typically to within 1 % of the Ferraris Wheel meter over a range of 0.1 to 1000 amps for any type of linear or non liner load.

This particular embodiment will produce results of power measurement which are theoretically in total agreement with the Ferraris Wheel meter when the supply voltage is an undistorted sinewave. When the mains voltage departs from this ideal shape the energy measured will still be theoretically extremely close for linear loads and remain very close for non linear loads.

Typically we expect an error of + .25% inaccuracy for linear loads and + 1 % maximum for non linear loads when the mains waveform has 5% third harmonic distortion.These figures for our electronic metering device are well within the accuracy as specified for the Ferraris Wheel electromechanical rotating disc meter.

As well as being capable of measuring true power VI cos f the embodiment of Fig. 7 can be used to measure reactive power VI sin + simply by sampling the current derived voltage at a point coincident with the voltage zero crossing time. Volt-amperes VI, can similarly be measured by sampling the current derived waveform at its peak. By appropriate changes to the processor's software the embodiment described can be arranged to measure one or all of the quantities VI sin +, VI cos , and VI.

Claims (18)

1. Apparatus whose measurement circuitry consist solely of electronic components and whose method of power measurement simulates the power measurement effected by a rotating disc electromechanical electricity energy consumption meter.

2. Electronic apparatus which simulates the energy measurement effected by a rotating disc electromechanical electricity energy meter by means of being sensitive only to the fundamental mains frequency component of load current and the fundamental mains frequency component of mains voltage.

3. Electronic apparatus which simulates the energy measurement effected by a rotating disc electromechanical electricity energy meter by means of being sensitive only to the fundamental frequency component of load current as well as being sensitive to the RMS value of mains voltage or the averaged value of a full or half-wave rectified derivative of mains voltage, or some other derived quantity directly related to the magnitude of the electrical supply voltage.

4. Electronic apparatus which simulates the energy measurement affected by a rotating disc electromechanical electricity energy meter by means of being sensitive to the instantaneous product of mains voltage and load current or deriving by process of voltage and current multiplication the same product while excluding any dc components of either quantity, and incorporating low pass filtering of the product variable with a filter characteristic simulating the response of the rotating disc meter.

5. Electronic apparatus which simulates the energy measurement effected by a rotating disc electromechanical electricity energy meter by taking into account load power factor by means of sampling the fundamental sinusoidal component of load current at a time halfway between zero crossings of the supply voltage waveform or at a time halfway between the zero crossings of the fundamental component of the supply voltage waveform.

6. Electronic apparatus which simulates the energy measurement effected by a rotating disc electromechanical electricity energy meter by means of sampling the fundamental sinusoidal component of load current at its peak, and sampling the fundamental component of mains voltage or the unfiltered mains voltage or a voltage linearly related thereto, and determining the phase difference between current and voltage waveforms by timing the difference between the corresponding zero-crossing times of current and voltage waveforms.

7. Apparatus according to claims 1 to 6 which incorporates an amplifier possessing a linear gain characteristic.

8. Apparatus according to claims 1 to 7 which incorporates an amplifier having a multiplicity of switchable values of gain.

9. Apparatus according to claims 1 to 7 which incorporates a number of values of gain which are selectable by the apparatus according to the magnitude of an input current signal, or fundamental component of such input current signal.

10. Apparatus according to claim 1 to 9 which incorporates an analogue to digital conversion device having a linear characteristic.

11. Apparatus according to claims 1 to 10 which effects integration of power to produce a totalized energy consumption indication.

1 2. Apparatus according to claims 1 to 11 which affects integration of power to produce pulses after each accumulation of a predetermined unit of energy.

1 3. Apparatus according to claims 1 to 12 which employs electrical, electronic, or magnetic sensing means to provide voltage or current inputs to the electronic circuitry whose amplitude varies in direct sympathy with the supply voltage and load current.

14. Apparatus according to claims 1 to 12 which employs electrical, electronic or magnetic sensing means to provide voltages or current inputs to the electronic circuitry whose amplitude varies in sympathy with the fundamental component of supply voltage or load current.

15. Apparatus according to claims 1 to 14 which effects measurement of power or energy consumption by means of sampling supply voltage and load current derived voltages once only in each cycle of the supply voltage or once only in a plurality of cycles of the mains voltage.

16. Apparatus according to claims 1 to 14 which effects measurement of power or energy consumption by means of sampling supply voltage and load current derived voltages once only or a plurality of times in each cycle of the supply voltage.

1 7. Apparatus according to claims 1 to 16 which effects measurement of power or energy where by suitable positioning of the single sample time or sample times within a single cycle of the mains voltage the apparatus is operable to measure one or several of the quantities of real power, reactive power or volt amperes passing to the load or the accumulated real energy, reactive energy, or volt ampere energy consumed by the load.

18. Apparatus for measuring electrical power or energy delivered to a load by means of electronic circuitry which simulates the power sensed or energy consumed as measured by a rotating disc electromechanical electricity energy meter as herein particularly described with reference to Figs. 1 to 7 of the drawings.