Bearings

11-99


Technical Feature

No power bearings spin energy from the wind

An active magnetic bearing which needs no power is the key to an innovative energy storage system for wind turbines. Tom Shelley reports

A new flywheel energy storage system has the potential to have low enough costs and high enough long term reliability to be used in remote locations and Third World countries

The flywheels rotate slowly enough to be used in only partial vacuum or hydrogen with very novel hybrid active magnetic bearings. These should in theory, require little or no power, and even at the present state of development require a mere 2W to suspend a fairly massive steel flywheel 250mm across.

Wind.jpg (28491 bytes)

Applications for the energy storage devices range from telecommunications to sewage treatment with the bearings offering solutions for machine tools and many other items of machinery requiring either high speeds or very smooth running.

Professor B. V. Jayawant at Sussex University has developed magnetic bearings for items ranging from sealed pumps for corrosive fluids and machine tools to mountings to silence an 8 tonne centrifuge used to produce concrete pipes. One of his former colleagues was the late Professor Eric Laithwaite, who shared the laboratory facilities until his death.

Professor Jayawant's latest development forms part of a JOULE-THERMIE project to find a better way of locally storing and smoothing energy supplied from 10 to 30kW wind turbines for powering such remotely located items as irrigation pumps, sewage treatment plants, refrigerators for storing medicines and telecommunications equipment. Rumour has it that telecomms relay stations are already using storage flywheels in the US, but the Professor hopes his designs will be cheap enough for use in his native India.

The project, which started in February, is being undertaken in conjunction with Proven Engineering; Nelco, who are making the static converter/inverter; Elettrorava in Turin, and the University of Patras in Greece. The budget is 800,000 ecus of three years, of which 500,000 have been contributed by the European Union.

At an early stage, Professor Jayawant says that the evaluating committee decided that they preferred the idea of flywheel energy storage to batteries because of the environmental problem of disposal of rechargeable batteries, and their relatively short working life. Design life of the average lead acid battery is about two years, compared with about ten years between major repairs and overhauls for wind turbines. Nelco's Geoff Humphrey, on the other hand, says that with frictionless magnetic bearings, the design life is at least 30 years, and since there is no physical rubbing contact, there is no real reason why they should not last for ever!

The key to the designs is to suspend them using the magnetic attraction between a magnetised cylinder and a circular steel plate. Since magnetic attraction increases as the gap decreases, the active part of the bearing is used to reduce and increase the overall magnetic attraction so that the weight of the cylinder, if suspended beneath the plate, is exactly balanced by the attractive force. Once the point is reached at which weight is exactly balanced by passive magnetic force, as produced by the magnets in the cylinder, it is only necessary to pass current through the active electromagnet coils to maintain the correct relative positions.

The amount of actual current required is vanishingly small, only sufficient to compensate for perturbations produced as the flywheel rotates, caused by irregularities in the components.

In the prototype flywheel, where the cylinder flywheel and motor generator are suspended beneath the plate, only an average of 1 amp has to be passed through the electromagnetic winding in the plate, the resistance of which is 2 ohms. Overall power consumption is thus 2W. With more accurate manufacture of components, it is hoped to bring the current down to 100mA or less, with an overall power consumption of 0.2W.

The cylinder/flywheel is constructed in the form of two mild steel cylinders, one inside the other. Between the two is a layer of ferrite loaded flexible plastic, identical to the material used for fridge magnets and seals. This on its own is not a very strongly magnetic material, but as Eureka was shown, it can develop a large amount of magnetic attractive force if a sheet is sandwiched between two sheets of steel, and the edge is used to attract a keeper completing the magnetic circuit. Even through a 3mm spacer, magnetic attraction is more than sufficient to support the weight of steel and magnetic elements.

The first small prototype was made using cut off sections of steel pipe and sections of magnetic sheet, none too neatly arranged, but works well enough.

The eventual goal is to build a flywheel and generator combination 1m across, rotating at 6,000 rpm, and weighing 100kg. The aim is then to produce larger flywheels capable of storing 2kWh or even more (10MJ) of energy at similar speeds.

Very fast running flywheels, as favoured in other research programmes, can store more energy per unit mass and volume, but require use of exotic materials such as aramid fibre to hold them together against centrifugal force. They also have to be run in a high vacuum since the peripheral speed is often faster than the speed of sound. Professor Jayawant's design, on the other hand, uses mild steel and runs in air. Air friction can be substantially reduced by placing it in an enclosure which can be partially evacuated or, in later versions, filled with hydrogen as is the case with the majority of conventional power station generators.

The 250mm prototype is mounted in a wooden enclosure. A steel framework to support the top cap was found to cause eddy current losses. Future structures are to be made of polypropylene. The enclosure is evacuated using a tyre pump, and the flywheel takes about three hours to run down from 1,800 rpm when the power is turned off. The final version is expected to lose only about half of its kinetic energy in six hours.

Many problems still remain to be resolved. The University of Patras, for example, has been set the task of devising a neural network controller which will coordinate the power output of the wind turbine and solar cells, and direct the appropriate parts of the electrical energy generated to an array of five flywheel motor generators running in parallel. At Sussex, they have already solved the problem of spinning up through the natural frequency of the bearing, which is about 5Hz, by changing the stiffness of the suspension. None of the remaining problems are thought to be insurmountable, and all the parties involved have a sufficiently well established track record to guarantee a successful outcome.

Low cost frictionless hybrid magnetic bearings will probably find their way even into household appliances in the not too distant future.

University of Sussex b.v.jayawant@sussex.ac.uk

Nelco nelco@polaron-group.co.uk

Design Pointers

Hybrid bearings used for the project employ a combination of passive magnetic attraction and a very small element of active electromagnetic force to ensure stability

The flywheels are likely to achieve five times the energy density of conventional lead acid batteries

Design lifetimes of the flywheels is expected to be 30 years, although barring externally caused mishaps, there is no reason why they should not last indefinitely.

Design pointers

Hybrid bearings used for the project employ a combination of passive magnetic attraction and a very small element of active electromagnetic force to ensure stability

The flywheels are likely to achieve five times the energy density of conventional lead acid batteries

Design lifetimes of the flywheels is expected to be 30 years, although barring externally caused mishaps, there is no reason why they should not last indefinitely.

Back to Bearings