GB2095487A - Induction Generators - Google Patents

Abstract

An electrical generating apparatus to be driven by a power source, which may use naturally occurring energy such as the wind, and to be connected to an a.c. electrical distribution system (20) includes a slip-ring induction machine (10), means for current transfer in both directions between the rotor (12) and the mains through inverter and rectifier systems (24, 26), and control means (28, 60) to determine the direction and magnitude of current transfer having regard to a set phase relationship between the current and rotor e.m.f. determined by phase comparator (30) and rotor speed dependent controller (60). The current transfer means may be semi-conductor switches. The machine frame size, transfer means rating and power source utilisation may be optimised by selection of a rotor slip speed range extending through system synchronous speed. <IMAGE>

Description

SPECIFICATION Induction Generators The invention relates to induction generators and control systems therefor to enable operation at a constant output frequency under conditions of variable rotor speed, particularly for applications in which the generator is driven by wind-power.

The possible use of wind-driven generators to supply power into the national electricity distribution system (the grid) is an aspect of energy conservation which is of increasing interest.

The problem of accommodating the natural variations in wind-speed has however not so far been solved in a way which is both technically and economically fully satisfactory. It has been proposed to use a wound-rotor induction machine in which at supersynchronous speeds a circuit connected to the slip rings is arranged to recover slip-power into the grid via a static frequency converter. A double output is thus provided from the stator and from the rotor.

It is an object of the invention to provide controlled machine operation over a relatively extended range which embraces sub-synchronous and super-synchronous speeds.

In accordance with the invention there is provided generating apparatus for connection to the electricity distribution system comprising a wound rotor induction machine, means for the mechanical coupling of the rotor to a power source and control means operative in response to the rotor speed to cause a flow current between the rotor and the distribution system in predetermined phase relationship with the rotor e.m.f. such that the machine operates with constant power factor, the value of the current being in predetermined relationship with the instantaneous speed of the rotor whereby the load torque of the machine in continuously matched to the driving torque of the power source over the range of speed variation.

The phase relationship may be synchronised so that the power factor is unity.

Preferably the phase relationship is predetermined by reference to a signal derived by comparing the supply frequency and the rate of rotation of the rotor.

The means for mechanical coupling may include gearing to enable a preferred range of speed at the source to correspond to a preferred range of rotor speed.

Current may be caused to flow in either direction such that the machine is controlled over a range of speed from above the synchronous value to substantially below the synchronous value.

The control means preferably includes electronic switches for the directional control of current flow. Such switches may be arranged as an inverter or as a cycloconvertor.

The limits of the range of speed can be selected so that the current and voltage ratings of the electronic switches are economic in relation to the power extracted from the source within that range.

The inverter has appreciated that, particularly for a wind generator, the induction machine must be matched to the source over the widest possible range of speed in order to recover economic quantities of power and that such matching requires a controlled current flow in to the rotor at sub-synchronous speeds as well as an outward flow at supersynchronous speeds. The current flow must be controlled both in value and in phase relationship with the rotor e.m.f.

Provided this relationship is maintained constant over the range of speed variation the machine is always operated at a constant power factor and the load torque is therefore proportional to rotor current. Usually the power factor will be held close to unity. For a windmill it is known that the driving torque can be made approximately proportional to the square of the shaft speed and thence torque matching is achieved by making the rotor current also proportional to the square of the shaft speed. A corresponding relationship can be derived for a power source having any different characteristics. The required precision of phase control is not possible by reference to the e.m.f. at the rotor slip-rings because contact noise predominates in the critical region near synchronous speed where the e.m.f. is close to zero.The present inventor has previously proposed in the context of the speed control of wound-rotor induction motors a means of deriving a precise indication of slip frequency and phase but the control of such motors represents a specialised branch of development in which the problems to be solved are different from those of a generator driven by a naturally fluctuating power source.

An embodiment of apparatus and its operation in accordance with the invention will now be described with reference to the accompanying drawings in which: Figure 1 represents schematically a wind-driven controlled generating apparatus in accordance with the invention; Figure 2 is an outline diagram of current control apparatus for use in the rotor circuit of the generating apparatus; Figure 3 illustrates the flow of current in the circuit of Fig. 2; Figure 4 represents schernatically apparatus for generating a control signal for use in the circuit of Fig 2; Figure 5 shows the derivation of the control signal of Fig. 4; Figure 6 illustrates the phase related waveforms of the control signal and the rotor e.m.f.;; Figure 7(a), (b) and (c) are graphs showing the variation of power and torque with the gearing between a windmill and a machine in the appartus of Fig. 1; Figure 8 illustrates the dependence of rotor circuit VA rating on gearing; and Figure 9 illustrates the dependence on gearing of the operation of the apparatus of Fig. 1.

With reference to Fig. 1 a 3-phase wound-rotor induction generator 10 is mounted on a shaft 12 which is driven through a gearbox 14 by a windmill 16. The stator winding of the generator 10 is connected by lines 1 8 direct to a 3-phase supply 20 derived from the grid. The rotor winding of generator 10 is connected to the same supply by way of control means which determines the timing and value of current flow between the rotor winding and the supply.

Current flow takes place to or from the winding via slip-rings 22 which are carried by shaft 1 2 and the flow path includes a current source inverter 24 and a phase controlled rectifier 26. The operation of inverter 24 is contolled by a control unit 28 is response to the output signal from a phase reference generator 30.

The operation of the inverter 24 and rectifier 26 will be explained with reference to the circuit diagram of Fig. 2. The a.c. supply 20 is connected to rectifier 26 on lines 32, 34, 36 each of which is associated with respective pairs of thyristors S1,S4; S3,S6; and S5,S2. The first of each pair is connected to the inverter in a common path 38 and the second of each pair is connected to the inverter in a common path 40 of opposite polarity. In the diagram of inverter 24 the rotor windings as seen across the slip-ring connections 22 are represented as three sources of e.m.f. e1, e2 and e3 each with an associated series inductance L1, L2, L3.

Thyristors T1 to T6 are arranged to enable current flow through each winding in either direction between lines 38 and 40.

First consider the windmill 1 6 to be driving generator 1O at a sub-synchronous speed. If when e1 is positive, as indicated by the left-pointing arrow, control unit 28 is arranged to fire transistors T1 and T6, a current will flow in the path 32, S1, T1, Di, L1 (the e1 winding), D6, T6, S6, 34 in the direction opposed to e1. That io to say, power will flow from the a.c. supply to the rotor. In the supersynchronous condition however control unit 28 is arranged to fire thyristors T3 and T4 during the positive half-cycle of e1. Current then flows in the path 34, S3, T3, D3, L1, D4, T4, S4, 32 in the same direction as e1. Thus power will flow from the rotor to the a.c. supply.The direction of current I in the d.c. link 38, 40 of course remains unchanged while a reversal of direction is observed at the slip rings 22. In order to simulate a sinusodial current waveform thyristors T1 and T6 (for the sub-synchronous condition) are operated only during the peak 60 of the e, half cycle to produce a current level 21/3. This value is less than I because of the presence of the parallel path through L2 and L3 across the winding L1. In the first part of the same half-cycle thyristors T5 and T6 are switched on, but the path through the winding L1 now also includes winding L3, with a parallel path through L2. The current is thus half the peak value. Similarly in the final part of that half-cycle thyristors T1 and T2 are switched on.The path through winding L1 includes L2 with a parallel path through L3. Similar sequences can readily be constructed for the other phases.

The resultant inward current flow is shown in Figure 3a for sub-synchronous speeds and the outward current flow is shown in Fig. 3b for supersynchronous speeds. In each case the horizontal scale represents time and in the vertical direction the sinusoidal waveform indicates the phase reference provided by the rotor e.m.f. e1 and the stepped waveform indicates the magnitude of current flow. For each step in current at levels 1/3 or 21/3 the reference numbers of the thyristors then conducting are indicated.

Controlled operation of generator 10 in the desired manner requires that the current flow be controlled in value and that it should occur (for the usual case of unity power factor) in strict phase synchronism with the rotor e.m.f. For the latter purpose, control unit 28 which generates trigger signals for the thyristors T1-T6 is supplied with phase reference signals from generator 30. The organisation and operation of generator 30 is shown schematically in Fig. 4, the basic function being to generate a synthetic slip frequency signal which is equal to the machine slip frequency but does not depend on measurements at the slip rings. The signal is generated directly from the difference between stator energisation frequency and the rotor speed.A small mains-driven synchronous motor 42 carries a toothed wheel or other means of generating a pulse train at a frequency of sixty-four pulses per cycle of mains frequency and the pulse train is input to a pulse splitter 44. A second pulse train is derived from a toothed wheel or other means 45 carried by shaft 1 2 of generator 10, so that the pulse frequency is proportional to rotor speed, and is input to a pulse splitter 46. In this case there are sixtyfour pulses per cycle of secondary e.m.f. The pulse splitters 44, 45 are controlled by different pulse outputs from a fourphase clock 48 which divides the pulse-train received from a master clock 49. If the input to splitter 44 goes higher after a clock pulse has passed the output is pulsed low on the rising edge of the next clock pulse. The output pulse provides an UP count to an UP/DOWN counter 50. In the same way splitter 46 provides a DOWN count to counter 50. Because the clock pulses are phase shifted between the splitters, the counter 50 can never receive an UP input and a DOWN output simultaneously and therefore no counts are missed. Counter 50 provides a 6-bit output on lines 52 to each count representing an increment of approximately 5.6 of the slip frequency cycle. The counter output signal is decoded and applied to generate a three-phase square wave output from unit 54. Since the output signals from unit 54 are the required phase reference input signals to inverter control unit 28 they must be synchronised with the phase of the rotor e.m.f.Synchronisation can be brought about automatically during a start-up sequence (which is initiated by a start circuit 56 having an input to clock 48) by holding the output from counter 50 to zero until the start of a positive half-cycle is detected in an appropriate phase of the rotor e.m.f. Thereafter synchronism is maintained without further reference to the rotor e.m.f.

It is sometimes desirable to vary the power factor from unity and a phase advance input 58 and phase retard input 60 to respective pulse splitters 62 and 64 enable pulses to be input to the counter 50 which displace the phase reference signal from synchronism.

The pulse sequences described as occurring in generator 30 are shown in relation to a common scale of time in Fig. 5, as sequences (a) to (j) in which, for synchronism; (a) represents the output from master clock 49; (b) +1 to 4 are the outputs of four phase clock 48; (c) is the mains frequency pulse train from motor 42 (d) is the unused Q, output at splitter 44 in response to +3 and train (c); (e) is the UP count output Q2 at splitter 44; (f) is the rotor speed pulse train from wheel 45; (g) is the unused Q1 output at splitter 46 in response to +, and train (f); (h) is the DOWN output Q2 at splitter 46; (j) is the 6 bit output from counter 50 representing slip frequency.

The phase relationship between the three-phase square wave output from unit 54 and the rotor e.m.f. is shown in Fig. 6. It is to be noted particularly that each square wave R, Y or B remains accurately in phase with the e.m.f. of the associated rotor winding closely above and below synchronous speed and provides a smooth transition through the abrupt phase discontinuity which necessarily at synchronous speed.

Generator 30 thus produces signals in response to which control unit 28 enables current flow through the inverter-rectifier system 24, 26 which is in the correct phase and in the direction appropriate to rotor speed. The value of the current is controlled by varying the phase angle of conduction of the appropriate ones of the thyristors S1-S6 in rectifier 26 and must be related to conditions at the rotor. Stability is maintained by setting the current so that the drive torque produced by the windmill at any instant is exactly balanced by the electrical load torque. Since the phase of the rotor current has been determined by the intial synchronisation of unit 30 such that the operating power factor is close to unity, the load torque will be directly related to rotor current.The drive torque, as previously indicated, is a function of speed, typically (speed)2, and the current must therefore respond to a speed signal which varies in the same way.

A direct speed signal is conveniently taken from toothed wheel 45 on shaft 1 2 to a rectifier control unit 60 in which the appropriate function is generated to control the order of firing and the firing angle of rectifiers S1-S6 and thus to pass the required rotor current. The output characteristics of the windmill, the gear ratio and the desired range of speed control will all be taken into account in determining the proportionality factor with which control unit 60 responds to speed.

In order to enable these relationships to be illustrated the power flow in the generator system will fist be considered. If the shaft power from the windmill is Pg then for a fractional slip S the power Pr crossing the air-gap of the machine is Pr = Pg/l-s. At subsynchronous speeds Pr is greater than Pg and, ignoring losses, the rotor must be supplied with power P2 = S Pr, so that the net power supplied to the grid (Pr-P2) will balance the input Pg. At supersynchronous speed S is negative and Pr is less than Pg and power P2 is recovered in the rotor circuit. Again the net power supplied to the grid (Pr + P2) balances the input Pg.Considering a single phase of the machine, the turns ratio between stator and rotor is denoted by P, the rotor current by 12 and the stator voltage, winding resistance and power by V1, R1 and P, respectively. The phase of the referred secondary current 12 can be automatically controlled by generator 30 to be at any angle (180 + O with respect to V1 and thus, neglecting iron losses, P, = 3 Vrssl2 cos 9 Then Pr = 3 [V1pl2 cos + + (fl12)2R1] and this quantity represents the electrical torque in synchronous watts.As before on the shaft from the windmill Pg = (1-s) Pr where the slip S is considered to be positive for subsynchronous generation and negative for supersynchronous generation.

Thus Pg = 3 (1-s) [V,ssl2 cos f + (fllA2R1] (1) Equation (1) enables the value of current 12 to be determined which is required at a speed corresponding to slip S to load the windmill to its design power Pg at that slip. Since the current may be controlled to match a wide range of speed above and below synchronous speed the remaining design problem is to choose gearing such that the operative range of slip is best suited to the characteristic of the windmill. The maximum power of the windmill may be arranged to occur at any desired value of slip S, but the variation of power over the working range and the consequent requirements in machine size amd control circuit rating must be evaluated.

The power of a windmill is generally assumed to vary with the cube of wind speed and thence similarly with rotor speed or with slip. Thus shaft power Pg at a slip S is related to the maximum power at slip S, in the ratio.

[(1-s)/(1-s1)]3 (2) Correspondingly rotor circuit power P2 = s. Pr = S. Pg/(l-s) = s.(1-s)2/(1-s,)3 (3) and stator circuit power P1 = Pg/(l-s) = (1-s)2/(1-s,)3 (4) To consider the torque produced at the shaft by the windmill, a square law variation with speed is assumed.

The ratio of Tg at any slip S to the maximum torque is (1-s)2/(1-s,)3 (5) for the condition that maximum windmill power is arranged to occur at synchromous speed.

Curves for power and torque as a function of slip for a range of values of s, determined by the gear ratio between windmill and generator are shown in Figure 7(a), (b) and (c). Curves 1 to 5 in each Figure relate respectively to S, = 0.1, 0, - 0.1, - 0.2. - 0.3. The rotor power P2 is seen in Fig. 7a to peak at sub-synchronous speeds and differentiating equation (2) and equating to zero shows that this maximum occurs at a slip s = 0.33. The value of s, that will minimise the maximum value of P2 can be found by equating the values of P2 from equation (2) for the points s = 0.33 and s = s,.

Thus 0.33 (1-0.33)2 s, (1-s,)2 (1-sa)3 (1-s,)3 gives s, = - 0.12 which is close to curve 3 of Fig 7a.

Substituting this value of s, into equation (3) gives P2(max) = 0.1 p.u. approximately.

Thus if the gear ratio is chosen such that the maximum windmill power is developed at a slip of - 0.12 then a maximum power of 0.1 p.u. will flow in the rotor circuit if energy is recovered from the windmill over its entire speed range.

In Fig. 7(b) shaft power Pg is plotted from equation (2); in Fig. 7(c) stator power P, and shaft torque Tg which are numerically equal in the units defined are calculated from equations (4) and (5). It will be observed that the maximum shaft torque at the generator falls as the value of s, increases negatively. This is to be expected as the windmill power is being recovered at a higher shaft speed and so a smaller frame size machine can be used.

It has been shown that curve 3 of Fig. 7a approximates to an optimum operating condition for minimising the power in the rotor circuit but the rating of the power semiconductors will be determined by the maximum voltage and maximum current to be controlled rather than by the in phase volt-ampere product at each operating point.

It can be shown that for constant power factor the load current is directly proportional to the shaft torque and as previously described, the current can be controlled by the generator 30, to be in phase with the rotor voltage. Thus the primary power is at unity power factor if f is set to be zero. Fig. 8 shows a curve 70 for rotor e.m.f. which is linear with respect to slip and curves 72, 74 for rotor current for which s, = 0 and s, = 0.3 respectively.

Consider the gearing to be such that P9(max) occcurs when s, = 0 so that curve 72 applies, the rotor current for that condition being denoted by 12(0). For s = 0 the stator power, neglecting losses, is given by P1 = #3 V1 11(o) cos # where l,(o) represents the stator current when power Pg(max) is being generated at s1 = 0.

As s, = 0 then P1 = Pg so #3 V1 I1(0) cos # = Pg(max) If ss is the turns ratio N2/N1 and E0 is the open circuit rotor voltage at standstill then E0 = ssVa and 12(0) = 11(0)/ss So A E0 12(0) cos ss = Pg(max) To minimise the secondary current the generator 30 will normally be synchronised as described such that cos f- 1 So the secondary Va, #3 E0 12(0) = Pg(msx) = 1.0 p.u.

If E0 is defined as 1.0 p.u. rotor volts then at any other slip E2 = sE0. If 12(0) is defined as 1.0 p.u. rotor current then the current at any other slip s for a gear ratio generating maximum power at slip s, is given by (1-s) 12= 1.0 pu.

(1-s1)3 The magnitude of the rotor voltage can be minimised if the operating range of the windmill is restricted to s = + s, when the maximum rotor voltage will be + s1 X 1.0 p.u. The semiconductors in the secondary circuit can be rated in accordance with this value provided the mains supply to the phase controlled rectifier 26 is suitably matched by using a transformer or tapped connection to the stator winding. Maximum VA will then occur at s = + s, when generating maximum windmill power. Then (1-s1) 12(0) E2 = Si Eo and I2(0)= (1-s,)3 1-s1 So the secondary circuit VA at any #3 E2 12 s, = x 1.0 #3 E0 12(0) s1 = p.u.

i-si for absolute value of 51 > 0 The power returned to the a.c. supply, neglecting magnetising current, will normally be at, or near, unity power factor as determined by the initial synchronising of the current source to the rotor e.m.f. The power factor of the rotor circuit, however, will change as the line commutated rectifier phases back to match the rectified rotor voltage. If the a.c. supply voltage VaC is such that the rectifier is at or near full conduction when the rotor e.m.f. is at its maximum of s1 E0 then

where Vdc and IdC are the values on the d.c. link.

Then as Vdc IdC = #3.Vac.Iac cos #2 approximately, where +2 is the secondary power factor it follows that cos 2 = 1.0 approximately.

At any other slip s over the operating range s = + s, the d.c. link voltage will be reduced by s/s, and so the firing angle of the rectifier must change such that cos +2 also reduces by s/s, to maintain the above power balance equation.

s So cos f5 = - approximately.

S, The dependence of the generator frame size on the selected value of s, is simply expressed. If the 1.0 p.u. frame size is defined as that required when the windmill is geared to deliver its maximum power at s, = 0 then, for any value of s, 1 Frame Size = 1.0 p.u.

1 -s, It will be noted that since the generator torque at s = s, is 1 /(1-s,) p.u. and thw generator speed is (l-s,).n,, sA being negative, the torque will decrease and the speed increase as s, is chosen to be a greater value of slip.

The effect of the choice of gearing, and therefore the value of s,, has been discussed with respect to the operating power range, the semiconductor VA rating, and the machine frame size.

It has been assumed that the operating range of the windmill is + s, and that outside this range the equipment is disconnected.

The results are summarised in Fig. 9 which shows the minimum power of the windmill that can be returned to the grid. Thus if the gearing is chosen to give s, = - 0.3, i.e. maximum windmill power occurs at a super-synchronous slip of 0.3, then energy can be recovered from the windmill dowm to 16% of its maximum rating with a semiconductor VA rating equivalent to 23% of the maximum power of the windmill and a frame size 77% of that that would be required if the gearing gave s, + 0. If the designer wishes to reduce the semiconductor VA then s, must be reduced say to s, = - 0.2 when the frame size of the generator will increase and the minimum power at which energy can usefully be recovered to the grid will be 30% with a reduction in semiconductor VA rating to 0.17 p.u.Thus there is a trade off between the VA capability of the semicondcutors in the rotor circuit and the on-line operating range of the windmill.

The power flow and power factor of the system s, = - 0.3, asuming that the supply side of the semiconductor equipment is voltage matched to the rotor e.m.f. and that all copper and iron losses are neglected, will be as shown in the table below.

Windmill Stator Rotor TOTAL s Power % Power % Power % cos < b5 cos (p -0.3 100 +77 + 23 1.0 1.0 - 0.2 79 + 66 + 13 0.67 0.98 -0.1 61 +55 + 6 0.33 0.96 0 46 +46 0 0 1.0 + 0.1 33 + 37 - 4 -0.33 0.95 + 0.2 23 + 29 - 6 - 0.67 0.96 +0.3 16 to 23 - 7 -1.0 1.0 where + defines power into the grid.

It will now be apparent that in carrying out the invention it becomes possible to recover energy from the wind or other fluctuating source over a wider range of rotor speed than has previously been known. The extended range of operation above, below and through synchronous speed enables the VA rating of the control devices to be minimised for a selected range of power recovery and also enables the frame size of the generator to be reduced. It is also important that the operating power factor can be held constant, and usually close to unity, whereas in the absence of the control system a variation in speed is accompanied by a variation in power factor. When that occurs it is of course impossibie to use rotor current as a control variable for torque.

It will be apparent to those skilled in the art that the form of control system described by way of example may be modified while still providing control of the specified parameters. For example it is possible to provide static frequency conversion and current control by means of a cycloconvertor instead of the d.c. link inverter as described.

Claims (11)

1. A generating apparatus for connection to an a.c. electricity distribution system comprising a wound rotor induction machine, means for the mechanical coupling of the rotor to a power source and control means operative in response to the rotor speed to cause a flow of current between the rotor and the distribution system in predetermined phase relationship with the rotor e.m.f. such that the machine operates with constant power factor, the value of the current being in predetermined relationship with the instantaneous speed of the rotor whereby the load torque of the machine is continuously matched to the driving torque of the power source over the range of speed variation.

2. An apparatus according to Claim 1 in which the phase relationship between the rotor e.m.f. and the rotor-system current is synchronised whereby operation is at unity power factor.

3. An apparatus according to Claim 1 or Claim 2 including means to derive a signal representing the comparison of the system frequency and the rate of rotation of the rotor and means responsive to said signal to predetermine the phase relationship with reference to said comparison.

4. An apparatus according to Claim 1 or Claim 2 or Claim 3 in which the mechanical coupling means includes gearing to enable a preferred range of speed at the power source to correspond to a preferred range of rotor speed.

5. An apparatus according to any one of the preceding claims including means to selectively cause current to flow in either direction between the rotor and the system.

6. An apparatus according to Claim 5 in which the control means selects said directional current flow and control means controls the machine speed over a range from above the synchronous value, for the system frequency, to substantially below said value.

7. An apparatus according to Claim 5 or Claim 6 including electronic switches for said selection of current flow direction.

8. An apparatus according to Claim 7 in which said switches are controllable semi-conductor rectifiers.

9. An apparatus according to Claim 7 or Claim 8 in which said switches are arranged as one of an inverter and a cycloconverter.

1 0. An apparatus according to Claim 5 or any claim dependent thereon, in which the range of speed is limited to restrict the electrical ratings required from the means to cause selective directional current flow.

11. A slip-ring generating apparatus for operation by a power source subject to natural variations in speed and for connection to an a.c. electricity distribution system over a range of generator machine speeds extending through system synchronous speed including means to control the direction and magnitude of current flow to the generator rotor and the phase relationship of the inverter with the rotor e.m.f., the magnitude being controlled with reference to the rotor instantaneous speed and the torque/speed characteristic of the power source and the phase relationship being held constant to maintain the torque/current characteristic of the generator without reference to the instantaneous rotor e.m.f. waveform.

1 2. An apparatus according to the preceding claims in which the power source is a windmill and the machine is run at speeds up to a supersynchronous speed of S, and a supersynchronous speed of S2 and disconnected at speeds outside the range S2 to S1 and in which S1 is up - 0.4 and preferably between - 0.2 and - 0.3 and S2 is at least + 0.3.

1 3. A generating apparatus substantially as herein described with reference to the accompanying drawings.