Stator Voltage Control Of Induction Motors - Online Article


Stator Voltage Control of an Induction Motor is used generally for three purposes – (a) to control the speed of the motor (b) to control the starting and braking behaviour of the motor (c) to maintain optimum efficiency in the motor when the motor load varies over a large range. It is simple in hardware and reliable compared to the more complex Variable Frequency Drives as far as speed control application is concerned. However, it turns out to be a somewhat dissipative method of speed control and results in lowered efficiency and rotor over heating. Fundamental aspects of Stator Voltage Control aimed at the above three objectives is covered in this article.

Stator Voltage Control for Speed Control of Induction Motor

Assume that we have a fixed frequency variable magnitude pure sine wave source available. Torque at any particular slip is proportional to (Voltage)2 in an Induction Motor. Fig. 1 below shows the Torque-Speed curves of an Induction Motor at various voltages assuming sinusoidal voltage. Also included is the torque-speed curve of a typical fan load. Note that the torque-speed curves of the Induction Motor clearly indicates that it is a motor designed for a large running slip. Otherwise the curves would have been steeper than this around full load rated speed. It can be clearly seen that the speed of the fan can be varied more or less uniformly in the range of 80% to 30% of synchronous speed of the motor by varying the voltage between 100% to 30%.

Motor with a running slip in the range of 2% to 5%) the possible speed range in the case of a fan load with voltage control would have been lower than this. The motor will pull out at around 55% voltage or so. Moreover, even the available speed variation will be highly non-uniform with the voltage variation. Hence it is necessary that an Induction Motor intended for speed control applications using stator voltage control has to be designed with a higher running slip and hence lower full load efficiency. (Because higher full-load slip implies higher rotor resistance). Such motors are usually designed with a full load slip value of 12%. Obviously their rotor must be of special design in order to withstand the higher rotor losses developed in the rotor.

The range of speed control available by voltage control is a strong function of nature of load torque variation with load. With a constant torque load, the available speed range is more limited than in the case of a fan load and the motor pulls out at voltage levels closer to 100%. Thus fan loads which have low starting torque demand, pump loads with little or no static head component in the system curve, blower loads with small starting torque demand etc. are the loads suitable for speed control by voltage variation.

Variation of Stator Current and Efficiency in Stator Voltage Control

For simplicity the analysis to follow will neglect the magnetising current of the machine. Though the magnetising current can be as much as 50% of full load current at rated voltage, it comes down rapidly with voltage and hence the above assumption is reasonable.

Assuming sinusoidal quantities, the average torque produced by the stator field reacting with rotor current is given by the following proportionality:

Tm a (I2R)/s where Tm is the motor torque, R the rotor resistance ,I the rotor current and s is the slip. Neglecting the magnetising current and core loss current the stator current is proportional to rotor current. If the load torque is related to the square of motor speed as is approximately true for a fan load

TL a (1-s)2 where TL is the load torque. At steady state the motor torque and load torque will be equal. This results in the following proportionality for stator current.

I a {(1-s)Ö s}/Ö R . This function has a maximum at s=0.33.Thus for a true fan type load stator voltage control will result in a maximum current at 66.7% of synchronous speed.

The ratio of this maximum current at s=0.33 to the rated full load current of the motor will vary sharply with the rated full load slip. The ratio is 1.75 if full load slip is 0.05 and it is 1.25 if the full load slip is 0.12.This implies that the maximum rotor copper loss (and stator Copper loss) in the machine will take place when the voltage applied is such that the fan load runs with a slip of 0.33 and that this maximum loss will be closer (but higher) to the rated full load copper loss if the rated slip of the machine is higher than normal. Usually the motors for this service are designed with a full load slip of 0.12 and hence their current can go up to 25% higher than rated value as the speed of a fan load is varied by varying voltage. Similarly their copper losses can go to 50% more than the full load copper loss under a variable voltage-fan load context. Thus we need an inherently inefficient machine to start with and the machine operation gets more and more inefficient at lower speeds. Hence this kind of speed control is used only on fan type loads and when only about 60% to 100% speed range is needed. Even then a motor with high rotor resistance (i.e. with a running slip of about 12%) should be used. Ordinary Squirrel Cage motors will suffer from rotor over heating on stator voltage control and should not be used for such service.

The Stator Voltage Controller

The stator voltage is controlled in these speed control systems by means of a power electronic controller. Normally thyristors in phase control mode are used. Various connection schemes exist. However detailed investigations into various connections had established in early eighties that the six thyristor-unconnected neutral scheme is the best in terms of minimum r.m.s current requirement and harmonic injection. Here two thyristors in anti parallel are connected between the line and motor in a phase. If the motor is star connected the neutral is left unconnected. This scheme was proved to take only 8% more r.m.s current (due to harmonics and converter induced reactive power requirement) than a pure sine wave source at the maximum current slip value of s=0.33.All other possible thyristor connections take more than this. Hence this 6-thyristor scheme is almost invariably used to control the applied voltage to Induction Motors in speed control schemes. The control is exercised by changing the firing angle µ of thyristors.

The thyristor controller brings in two more sources of power loss. Power loss takes place in the power devices in the controller. In addition, harmonic losses take place in the motor due to harmonic currents flowing in the winding due to phase control. These two additional loss components will make this speed controller further inefficient. Also over heating of the motor on harmonic losses is another possibility. Harmonic currents can result in cogging/crawling etc. especially when attempts are made to run the motor at very low speeds.In spite of all these problems these speed controllers are popular ;especially for fan loads with limited speed variation, due to their simplicity and reliability.

Starting/Stopping Control by Stator Voltage Control

The commercially available Starting Torque Controllers (STC), Soft Starters etc. make use of stator voltage control using the 6-thyristor scheme to control the starting/stopping behaviour of the motors. They behave like a continuously variable autotransformer during starting. After starting, the thyristors are shorted by contactors to avoid device losses and full voltage is directly applied to the motor. More details on these controllers are provided in another accompanying article.

Stator Voltage Control for Optimum Efficiency Operation of Motor

Induction Motors are highly efficient at rated load and have efficiencies in the range 85%-95%. Motor losses consist of three main components: (1) Friction and Windage Losses (2) Iron Losses (3) Copper Losses. Friction and Windage Losses are insensitive to load changes, as speed is essentially constant.

Iron losses consist of hysterisis losses and eddy current losses. At constant frequency, hysterisis loss is proportional to B1.6 and the eddy current loss is proportional to B2 where B is the maximum flux density in the air gap. The maximum flux density remains constant if the applied voltage is kept constant. Thus, as load is decreased, voltage remaining constant, the iron loss constitutes a greater percentage of the output. This results in poor efficiency at part loads.

Part load efficiency can be improved by reducing the applied voltage to the motor. The motor has to be a standard squirrel cage motor optimised for full load running. In the case of such a motor the running slip will be around 0.04 and hence its torque-slip curve will be steep around zero slip. When the applied voltage is reduced, the load torque intersects the motor curve at a new point on the new torque-slip curve. However due to the steepness of T-s curves, the speed of machine will not vary much though it will decrease a little. Hence as a first approximation it may be assumed that the motor speed does not change when voltage across an underloaded motor is varied. If the speed does not change the load torque and mechanical power output will not change. And since voltage has come down the motor will draw an increased active current component to supply the same output. The reactive current component is predominantly magnetising in nature and it will come down since applied voltage has come down. The total stator current which is constituted by active and reactive components can increase or decrease depending on the amount of voltage reduction. Thus when the voltage across an underloaded motor is gradually reduced its stator current decreases first, reaches a minimum at a particular voltage and increases with further reduction in voltage. The value of minimum current will depend on the exact load on the motor.

Coming to the loss variations, with reduction in voltage the iron loss comes down. And initially the currents and hence copper losses also come down. When the voltage is reduced to sufficiently low level, the consequent increase in copper losses will at some point turn the total losses away from its decreasing trend i.e. there will be one particular voltage at which the total losses in the motor will be a minimum. This voltage value will not coincide with the voltage value at which the current is a minimum at the same loading level; but they will be close.

With the assumption that the speed of the motor does not vary with reduction in voltage the minimum current point will coincide with maximum power factor (or minimum phase angle) condition. Similarly the minimum loss point (i.e. maximum efficiency point) will coincide with minimum power input point. Minimum current point does not correspond to maximum efficiency point as already mentioned; but they are close. But if the small variation in motor speed and consequent changes in output power are also considered, the optimum voltage point for a particular load condition in the four cases i.e. the minimum current point, the minimum power factor angle point, the minimum power input point and the minimum loss (maximum efficiency) point, will be different. The minimum loss point is difficult to monitor electronically; though that is what we want to do. However, the other three conditions can be monitored electronically by sensing motor voltage and current and using some form of a minimum search algorithm implemented either digitally or in analog circuits. Of course, the loss reduction achieved will be less than optimal. Minimum power condition is the closest to maximum efficiency condition followed closely by current minimum condition. It is easier to process the current minimum search and hence it is current minimum search that is employed in most of the Smart Motor Controllers available in the market.

The six-thyristor scheme is used in all these SMCs. The SMCs also take care of the control of starting and stopping of the motor also. More details on SMCs are included in another accompanying lecture. Essentially, they start up the motor and apply full voltage first. Then, the current is sampled. A search routine is initiated. The voltage is decreased slightly and the change in current is noted. If the current decreases the voltage is further reduced in steps till the current shows a tendency to turn back i.e. to increase. If, in the first voltage reduction step the current increased, then the voltage is taken up in steps till the current reaches a minimum. This procedure is repeated in a periodic manner to fine-tune the applied voltage against load variations.

Possible Energy Savings by Using Smart Motor Controllers

The amount of energy savings possible depends on the extent of loading on the motor and its duty cycle. It also depends very much on the optimal operation algorithm implemented in the Controller. Indicative figures on possible energy savings arrived at by a Westinghouse Corporation Study is summarised in Figure 2.

Reconnecting a Motor in Star for Energy Savings

It was stated earlier that for a certain part load on the motor there is a particular voltage to be applied to a motor such that its losses will be a minimum and its efficiency would be a maximum. Obviously there is a particular part load percentage for which the optimum voltage is 230 V in a 400 V rated motor. For that load one does not need a controller; simply reconnecting the motor in star from delta will result in optimum loss. Usually this load is around 20%. However reconnecting the motor in star when its loading is between 5-30% will result in some energy savings at all loads in that range, though not optimal.

This reconnection is inexpensive and can be used if the loading percentage is less than 30% and the motor is always running with that loading or lesser loading. The overload relay has to be set at 60% of its original setting after reconnection. Also, if there was a capacitor connected across the motor when it was on delta it should be removed after reconnection. Experimental values for power savings in watts and energy savings as a % of energy consumed prior to reconnection are provided below for two motors. On a percentage basis roughly the same amount of savings can be expected on a motor of any rating at same loading level.

About the Author:

No further information.


No comment yet. Be the first to post a comment.