LARGE INDUSTRIAL DRIVES
Reproduced from Machinery Lloyd and Electrical Engineering (No. 5-17th July. 1965) with kind permission of the publishers – The Certificated Engineer November 1965.
In recent years the Institution of Electrical Engineers has gained an enviable reputation for the quality of its many symposia. There have been times, of course, when purely theoretical treatment of some subjects has made a paper of limited interest only and left the practical engineer in the audience more confused for his pains. No such criticism could be levelled at the recent conference on the Application of Large Industrial Drives held by the Power Division of the Institution, Savoy Place, London WC2.
Many of these drives have been developed within individual industries for their specialised applications, but as the basic requirements are often similar much fundamental development work has been needlessly duplicated in the past. Steel working, mining, chemical and public utilities are representative of the industries which may benefit from the exchange of information and indications of future developments given at such conferences.
One example of the lack of this type of communication between the industries was given when a speaker quoted a decision which had to be made regarding the use of an automatic transfer system in a nuclear power station. It was believed at the time, 1958, that such a scheme had never before been used in the UK and that no first-hand information about the prolonged operation was available. It appeared therefore that the decision would have to be made without this knowledge. Quite by chance, however, the authority concerned heard about the successful application of two such systems in the petrochemical industry. Collaboration between the different concerns ensued and the design was satisfactorily completed.
The Symposium
This was split into six conferences each of which had its theme and which collectively provided thirty-two papers.
The first conference, 'The effect on the Supply and Distribution of ac and dc Drive' consisted of five papers, four of which related to the design of electrical systems. These paper dealt with the problems facing the system design engineer and stressed the need for detailed studies of alternative systems to obtain optimum economic and operating conditions. The paper, by K. K. Schwarz, MA, MIEE, of Laurence Scott and Electromotors Ltd, entitled 'The Effect of Motor Choice on Supply System Design' indicated the conflicting problems that can arise. Mr Schwarz considered three basic effects of a large motor on the supply system: starting current, recovery current after short time interruption and harmonic generation. These parameters all affect the required short circuit level and therefore the supply transformers, circuit breakers, cables, and finally the voltage of the system.
Starting problems are associated mainly with squirrel cage motors where the conflicting demands of a high starting torque combined with a low starting current and a high running efficiency have to be met. Mr Schwarz put in a plea for starting requirements to be limited to those required by the drive.
If the supply system is temporarily interrupted or the voltage heavily depressed it is often essential for process reasons that the motors will recover, i.e. regain normal operating condition on the restoration of the supply. Two extreme general cases were mentioned. Low inertia constant torque drives result in rapid deceleration of the equipment, so that on recovery what is virtually a normal start Occurs. This usually requires the tarting equipment being returned to service so that continuity of output is lost anyway. The only solutions are to add mechanical inertia or to provide an alternative supply. High inertia falling torque drives, on the other hand, suffer very little deceleration in the 3 to 5 seconds usually considered and output is maintained.
Harmonic generation is a problem which arises in rectifier fed dc drives, static frequency changers and some induction motor cascade with static recovery equipment. A comprehensive series of tables included in the paper showed that this problem can often be more severe than the occasional starting current.
The author ended his paper with two realistic examples which illustrated the complex design requirements met within a boiler auxiliary system and a pumping station supply system. The paper concluded by claiming that stator fed variable-speed motors are found to have advantageous properties in several respects. Consequently, they may result in an overall economic advantage in operation.
The fifth paper in the first conference gave the actual results of a study on the starting and stability of large squirrel cage induction and salient pole synchronous motors of 22 500 h.p. when connected to an existing 11 kV industrial distribution system and direct-on-line started.
Entitled 'Machine Design to Suit the Application', the second conference in the symposium consisted of six papers. Two of the papers were complementary in their treatment of torsional oscillations in a rotating machine. The authors of both of these papers used a computer in solving some of the calculations involved.
The effect of transient current and torques was considered by one paper with particular regard to induction motors during switching operations. These transients may reach several times steady-state values and may be an important factor in the design, as the first peak of transient torque may be opposed to the direction of rotation of the machine, and pulsations of torque with an amplitude greater than the full-load torque may persist for many cycles of the supply, particularly with a lightly damped machine.
One of the remaining paper outlined, by giving typical performance figures, the limits of synchronous and asynchronous machines. The authors of this paper, Mr G. Cambell, BSC(ENG), and Mr P. H. Conuica, BE, of the General Electric Co Ltd, concluded that it is unwise to reduce the torque margin too far unless the motor performance and supply system characteristics can be accurately forecast under all conditions. They recommended the retention of an adequate torque margin, during acceleration at least 10 per cent (and preferably 15 per cent) greater than the load torque under the worst estimated supply system conditions. Drives Employed in Rolling Mills
The application and control of drive in the metal processing industry received considerable attention from the paper in different sections of this symposium. Tn section 2 the paper by N. G. Lindholm of Asea, Vasteras, Sweden. entitled 'Special Feature in the Design of Reversible Rolling Mill Motors in Sweden', described the design of a de motor which was developed to overcome the usual problems of inertia and overload torque associated with such drives.
The machine uses spokes to transmit the torque from the armature core to the hub. which allows a considerable having in weight and results in the desired low moment of inertia without resulting in an unduly long machine of small diameter with the consequent commutation difficulties. The spider-type construction also aids ventilation. The other main advantage of this spoked construction is that it is possible to use a hub with variable strength. In this way a more uniform transmission of torque along the entire hub is obtained, thus reducing local tresses.
Methods of armature winding were also described for obtaining satisfactory commutation over the wide speed range covered by this type of drive. The splitting up and transporting of armature conductors to reduce eddy currents was mentioned.
So far with reversible motors, it has been the usual practice to use radial brushes. These have several inherent disadvantages, however, not the lea t of which is the extreme accuracy required in setting the brush nose positions. A new type of holder was described which overcomes many of these disadvantages. Designed for use on large reversible motors, this brush holder has two separate boxes with a pressure finger for each. The construction of the brush-hand holder is such that it always has it large face against the wall of the brush box. At normal values of the coefficient of friction, the brush shows no tendency to tip over and the angle of the holder to the commutator is not critical. An error of a few degrees is allowable. To improve commutation still, further, the individual brushes are split.
Mr D. A. Kennedy, AMIEE. of English Electric Ltd. also dealt with dc rolling mill motors. His approach was to list seven of the most important, and sometimes conflicting, pressures and then to expand the design features which result. For example, on the subject of rotor inertia, Mr Kennedy made the point that length is not a complete guide to reliability and material should be used most efficiently. More specifically, it could be shown that machines with wide speed ranges obtained by field weakening will be most likely to be limited by commutation, whereas for machines with small speed ranges the minimum diameter will be determined by the maximum permissible voltage per segment.
Mr Kennedy had much to say on reliability, ease of maintenance and the foolishness of awarding contracts on the lowest price basis, As the third section was concerned with power conversion system and control and regulating systems such as speed variation it was not surprising that five of the 10 paper in it dealt with rolling mill drives. three on the general application and two on control. 'Control of large Reversing Rolling Mill Drives' was the title of the paper by J. McTaggart, BSC, AMIEE, of Associated Electrical Industries Ltd, Rugby. It illustrated detailed control design and gave an example of a complex dc control circuit. An eros connected converter was considered. For simplicity, it was assumed that there were full voltage and full field on the motor, the controlled variable being motor peed. However, to allow for the varying operating condition it was necessary for the no-load speeds of the two motors concerned to be suitably mismatched. When the piece entered the rollers, the relative roll speeds were fixed and it was then required to share the load between the motors. A load balance circuit was used to achieve this, which compared motor currents and altered the speed control loop on one or both motors. Should one of the work rolls slip the load balance signal should have enough damping to prevent overvoltage or over peed. Each motor had a speed regulator, which consisted of an amplifier designed to compare the actual speed with the required speed as dictated by the load balance signal and sets the demand for a current regulator. This regulator modified the grid firing angle, of the converter, in the appropriate direction to maintain equality. By limiting the output of the speed regulator the maximum motor current was readily held within the permitted maximum. Ala, by inserting a ramp unit between the current limit circuit and the speed regulator output it was possible to keep the maximum rate of change of current to a safe value. In this type of cross-connected converter system, each converter passes current in a different direction according to its polarity. Control was obtained by a current regulator, for each converter, to which the current demand signal was directed according to it. sign. Also, there was a signal element representing the circulating current.
Typical response times for small disturbances were 20 msec for the current regulator and 100 msec for the speed regulator.
The paper concluded by stating that the relay equipment used for the present-day logic functions, associated with these drives, is being replaced by static switching circuits. These have increased the speed of response. Converter powered drives at present use controlled rectifiers, mercury arc or thyristors to supply the motor armature. In motor-generator gets they are used to supply the motor and generator fields.
Comparison between motor-generator set and converter powered drives is mentioned in the paper by Mr A. R. Carrie, BSC, entitled 'Effect of Steel Rolling Mill Drives on the Power Supply'. A motor-generator set rated at 12 000 hp can demand 27.5 MVA from the supply following the sudden application of load. This could be imposed in just under I sec on a motor-generator set. The converter equivalent would demand a supply swing possibly 10 per cent greater although over a shorter period.
Harmonic generation is an inherent problem when using converters and it is customary to find main drive converters with a 12 phase input to reduce the output de ripple. As a result, 11th, 13th, 23rd and 25th harmonics are predominantly present in the supply. A possible solution is suggested in the phasing of groups of converters apart. If two main drives are liable to peak at the same time then their converter transformer secondaries phased 15° apart, with 12 phase grouping on each, would simulate a 24 phase grouping. Filters must be used if the harmonics present are of a magnitude not accepted by the electricity authorities.
Site Operation
Operating conditions, safety and maintenance were covered in the fourth conference of the symposium. Preventive maintenance was featured in several papers and the need for its acceptance as a useful engineering tool by the user. On the other hand, manufacturers were urged to make machine and equipment more readily accessible. Unit construction was mentioned as a desirable feature so that faulty parts may be removed for repair and healthy components substituted. Comprehensive monitoring would also be an asset as it would allow the normal standard of maintenance personnel to be engaged in fault finding. This is important as many of today's larger drives posse complex ancillary equipment, which can only be fully serviced by relatively highly qualified technicians.
One of the papers which referred to site operation was that by J. R. Hazel, AMIEE, of the Central Electricity Generating Board. The paper covered subject matter under the heading of 'Large Motor Drives in Power Stations'. The largest motor in a modern thermal generating station is the boiler feed pump drive motor. This single item of equipment has a profound effect on the auxiliary distribution system and in the large stations now being built the motor for this application would have to be rated at about 10 000-15 000 hp. However, in these cases, the task is handled by a steam turbine is driven boiler feed pump. which uses team bled off at an intermediate stage in the main turbine.
Among the advantages claimed by this system over the electrical equivalent are: higher efficiency, easier speed control and a lessening of the load on auxiliary distribution systems.
Electrically driven feed pumps are used for starting and standby duties. Such motors are typically rated at 7 900 hp at 11 kV for a 550 MW unit and are normally two-pole machines. Smaller motors of about 1 000 to 2 500 hp are used for driving water pumps, draught fans and condensate pumps. It is not unusual to find a modern power station with 20 associated motors of 1 000 hp or over.
Most of these machines are of the squirrel cage induction type. For induced draught fan applications they are of two-speed design and usually have two winding, each with a separate terminal box. A usual design would be an 8/10 pole machine.
Pole amplitude modulation motors are now available (Machinery Lloyd, European Edition April 1965) which fulfil many of the requirements of the two-speed drives and some of these machines are being constructed for this purpose.
Mr Hazel's paper gave an instance of a recent two-speed installation which illustrated a typical specification for an induced draught fan drive on a modern power station: 2 140/1 090 hp, 744/596 rev/min, 6.6 kV, CACW squirrel cage motor. The moment of inertia of each fan was 160 000 Ib ft2 and acceleration time with direct-on-line starting was 45 sec. Switching was using an oil circuit breaker.
Outputs of the order of several thousand hp have also been obtained from ac commutator motors, at 3.3 kV, on variable speed drives such as fans associated with coal milling plant.
Remote operation of the large motors discussed usually requires a comprehensive system of monitoring devices and alarm indicators. One of these machines may be equipped with upwards of 15 alarm and monitoring devices. Included in the operating parameters which are usually metered are bearing temperature, oil pressure, winding temperatures, cooling water flow and running current. The number of stars which can be safely made within a given time is another vital aspect in the operation of large induction motors. This problem arises because, during the starting period, these motors have to absorb a large amount of energy. The rotor has to be designed so that its thermal capacity is not exceeded during the run-up. Damage can also be caused to end rings, bars and joints by too rapid expansion. A limit has therefore got to be set. Normally at least two successive starts are specified.
The last two sessions of the symposium were the shortest. The penultimate conference dealt with the selection of motors for particular purposes and the factors involved in each case. On the economic side capital costs and running costs were considered as well as power factor correction as a means of reducing the latter. The paper, by P. F. Grove. MIEE. of McLellan and Partners dealt with this subject and provided a relationship for calculating the optimum pf for a system. Equations were also given for determining the time taken for the return of capital after installing a pf correction scheme.
The remaining papers in this section covered specialist applications including colliery winder drives, a foul air fan drive and the operation of squirrel cage motors in carbon dioxide at 20 atmospheres pressure.
Symposia such as this usually provide the participants with plenty of food for thought and the desire to increase the frequency of such meetings. Under the heading, 'Developments and Future Trends' the la t conference in this symposium certainly did this.
Pole amplitude modulation was again mentioned in this section in a paper headed 'Multispeed Induction Motors for Boiler Draught Plant', by H. Sterling, AMJEE, Lancashire Dynamo and Crypto Ltd. It is claimed that for some of the above applications the pam motor is smaller, cheaper and more efficient than its dual wound equivalent. The problem of more complex control gear associated with these machines has been eased by the development of an off-load changeover witch operating in conjunction with a circuit breaker.
The second paper dealt with studies undertaken to find out the largest direct-on-line tarted motor that could be used on power station auxiliary drives. The paper also mentioned induction motor contribution to faults-such a motor act as a generator during the first few cycles after a short circuit occurs. The author, Mr D. I. Thornton, MA, of Kennedy & Dukin Ltd, concludes that direct-on-line started motors up to about 11 500 hp can he u ed with II kV, 750 MVA breaking 900 MV A making switchgear without the voltage drop on starting was 15 per cent. Above this direct-on-line started motors up to 20 000 hp can be used under the same conditions if an automatic live changeover scheme is adopted, and above 20 000 hp resistance started motors or motors connected to a higher voltage could be used. Of the remaining papers, one covered the use of polyphase thyristor inverters. Variable frequency schemes, using these devices, are now a practical proposition for feeding induction motors on variable speed drives. There is little doubt that such systems will become a close rival of the dc motor and many of the applications handled by the latter, such as roIling mill drives and traction, will be suitable for the inverter-motor combination.
The last paper in the symposium reinforces the argument for synthetic frequency variation by stating that synchronous machines are ideally suitable as they could provide synchronous torque at all speeds without the usual starting troubles. The paper covered a wide range of present problems associated with various types of motors and indicated future developments to overcome these problems. The potential of the above-mentioned thyristor systems was: stressed-slip energy recovery and replacement of commutators on de machines. The use of computers was also mentioned as an aid to design and they were used by some of the authors in compiling results for their papers.