Willi Paul, Makan Chen, Werner Hofbauer, Michael Mendik
Figure 1. ABB’s expertise has produced the world’s most powerful superconducting fault current limiter
Developments in high temperature superconductor materials are bringing superconducting fault current limiters closer to practical reality. These high temperature superconductors will prevent damage to power systems and networks from short circuits.
Electricity companies invest heavily to protect their power systems and networks against the risk of damage by a short circuit. Ageing insulation, an accident or a lightning strike can unleash a massive fault current, which is limited only by the impedance of the system between the fault location and the power source. In the worst case, the short-circuit current can be around one hundred times the nominal current under normal load, leading to high mechanical and thermal stresses in proportion to the square of the current value. The situation is made even more serious by the ever increasing world-wide demand for power which has led to the growing interconnection of electric grids, resulting in increased short-circuit currents.
Figure 2. The conductivity of a superconductor can assume has three states: zero resistance, inside the innermost region; a transition region; and normal conductance
All the components in an electrical power system have to be designed to withstand these additional short-circuit stresses for a certain period of time, which is usually determined by the interval before the circuit breakers act to interrupt the fault current, and varies typically between 20 to 300 milli-seconds. The higher the anticipated fault current then the more the equipment costs, and the higher fault currents also increase maintenance costs since they accelerate ageing.
Naturally, any device which can reduce the prospective fault current in the system will enable its components to be down-rated, leading to significant cost-savings. Hence the demand for fault current limiters (FCLs) to act as effective ‘shock absorbers’ in order to limit the flow of power in an emergency without hindering its flow during normal operation. Most FCL concepts depend on either mechanical means, the detuning of inductance-capacitance (LC) resonance circuits, or on components with strong non-linear behaviour (eg: semiconductors, iron-core reactors, superconductors). They all have some practical drawbacks. However, among the non-linear materials, superconductors appear tailor made for use in fault current limiters because of their unique, sharp transition from zero resistance at normal currents to a finite resistance at higher current densities.
Superconductors were first discovered in 1911 when a Dutch scientist, Heike Kammerlingh Onnes, cooled mercury down to -269°:C. It was found that below certain critical values of temperature, magnetic field and current density, superconductors would lose all their electrical resistance.
In the innermost region of a superconductor, where the temperature, field and current density are low enough, the material is in its true superconducting state and has zero resistance. In the next region, the resistance increases steeply as values for the three variables go higher. Outside that area the resistance is effectively independent of field and current density as the material reverts to behaving as an ordinary conductor. See Figure 2.
There are essentially two types of superconducting fault current limiters (SCFCL), namely the ‘resistive’ and the ‘shielded core’ or ‘inductive’ concept.
The most straightforward SCFCL concept is the resistive one, in which the superconductor is directly connected in series with the line to be protected. To maintain its superconductivity, it is usually immersed in a coolant that is chilled by a refrigerant. In normal operation, the current and magnetic field can vary, but the temperature is held constant. The cross-section of the superconductor is designed to enable it to stay below the critical current density. In this regime the resistance is zero which means the impedance of the SCFCL is negligible, and does not interfere with the network. However, the superconductor’s impedance is only truly zero for dc currents.
Figure 3. The most straightforward SCFCL concept is the resitive one, in which the superconductor is directly connected in series with the line to be protected
In the more common ac applications two factors apply. First, the finite length of the conductor produces a finite reactance, although this can be kept to a minimum by special conductor architecture. Second, a superconductor is not loss free; in an ac operation the alternating magnetic field generated by the current produces ac losses, which are basically eddy current losses. These losses are heavily influenced by the conductor’s geometry and can be reduced by decreasing its dimensions transverse to the direction of the local magnetic field. They contribute little to the total impedance of the SCFCL, but they do dissipate energy in the superconductor, raising cooling costs.
Should a fault current arise, the inrush of current and magnetic field will push the superconductor from the innermost region of Figure 2 into the transition region, between zero resistance and normal conductor behaviour. This sharp growth in the resistance causes a first limitation of the of the fault current. The following temperature rise leads to a further amplification of the resistance and thus to a further reduction of the current, before it is interrupted by the breaker.
The shielded core SCFCL (also sometimes known as an inductive SCFCL) is basically a shorted transformer, where the superconductor is magnetically connected to the line to be protected. The primary coil of the device is a normal conductor connected in series to the line to be protected, while the secondary side is superconducting and shorted. In normal operation the iron core sees no magnetic field because it is completely shielded by the superconductor – hence the name ‘shielded core’. Assuming an ideal transformer, the shielded SCFCL will show the same behaviour as a resistive SCFCL. Since the number of turns of the secondary winding may be far fewer than on the primary, only short superconductors are needed. In fact, in most approaches there is just one secondary turn, with the superconductor winding in the form of a tube.
The technical feasibility of using low temperature superconductors (LTS), which are mainly metals, alloys and intermetallics, as the basis for prototype superconducting fault current limiters was established by groups such as GEC Alsthom and Electricité de France (EDF), and Toshiba and Tepco. However, they proved impractical as the LTS needs to be operated at extremely low temperatures resulting in high cooling costs.
The situation changed dramatically in 1986 with the discovery of high temperature superconductors (HTS), which can be operated at much higher temperatures (-196°:C) and can be cooled by liquid nitrogen. However, the construction of the long conductors required for SCFCL applications has been a significant challenge since the ceramic material is very brittle.
Figure 4. A shielded core or inductive SCFCL is a shortened transformer where the superconductor is magnetically connected to the line to be protected
ABB’s developments in SCFCLs has been based on the use of the BSCCO type (known as ‘Bisco’) bismuth based high temperature superconductors, specifically Bi2212. In 1996 ABB installed its first prototype device, with a rating of 1.2 MVA, in a Swiss hydropower plant (NOK). This device utilized the shielded core design, principally because it requires smaller lengths of the HTS material, and proved the feasibility of the technology over a year-long test period.
A new design
The disadvantage of the shielded core design is its relative complexity as well as its large size and weight. ABB’s recent efforts have focused on the simpler and more compact resistive design, but this requires much longer continuous lengths of the HTS material. The breakthrough came with the development of a novel manufacturing technology that now enables these long lengths to be produced more cost-effectively. Instead of using wire windings, the ABB system is based on large sheets of the Bi2212 ceramic material mounted on a supporting metallic base that protects it from cracking. This sheet, which is produced specifically for use in SCFCL applications, is around 0.25 m2 and 3 mm thick. In order to duplicate the effect of 200 m of wire ABB uses lasers or water jets to carve a long series of meandering channels onto the metal-ceramic sheet.
Using this flat structure approach has enabled the volume of the SCFCL device to be reduced by a factor of five, compared with the previous 1.2 MVA prototype, resulting in a core system of less than 50 kg. The first prototype of the new resistive design, which was single phase, had a nominal current of 200 A[rms] and was tested in circuits with nominal voltages up to 8.3 kV[rms], corresponding to a rated power of 1.6 MVA. Over a programme of 20 tests the device limited both symmetrical and asymmetrical fault currents without degradation of the HTS components. At 8.3 kV[rms], the SCFCL limited a prospective fault current of 13.2 kA[rms] to a first peak of 4.3 kA. Heating of the material further limited the fault current to 1 kA[rms] after 50 milliseconds.
The latest prototype based on this design has recently been tested to a rated power of 6.4 MVA, which makes it the world’s most powerful SCFCL based on HTS to date.
The key practical advantage of the SCFCL is that it will start to limit any fault current as soon as the critical current of the superconductor component is exceeded, because its resistance increases rapidly outside the superconducting state. This critical current is adjustable between a few and several kAs by varying the cross-section of the superconductor and it does not need any mechanism to switch it into the higher resistive state. Because the switching is based on the intrinsic physical property of the superconductor it is inherently fail-safe, which means that the SCFCL cannot fail to limit the current.
Figure 5. Evolution of current and voltage over the 1.6 MVA SCFCL during short-circuit testing. Reproduced by kind permission of Cigre
In addition, the physics of the limitation process is absolutely reversible, in contrast to a circuit breaker operation, which always leads to an irreversible contact ‘burn-off’. Superconducting current limitation is a solid state process, there is no arc involved, so the possible number of limitation processes is infinite.
Many modern industrial processes and sensitive microelectronics call for premium power quality. Yet at the same time the universal use of power electronics is leading to increased grid pollution such as harmonic and voltage dips (flicker), while arc-welding machines and arc furnaces also contribute to disturbances. Currently, the most common solution is to connect these customers to a very strong grid (ie: one with a high short-circuit power), which would be the grid of the next higher voltage level. In most cases this would represent a very costly solution.
Figure 6. Possible MV applications for SCFCL – values in brackets can be achieved if SCFCL is installed. Reproduced by kind permission of Cigre
The use of an SCFCL solves the old contradiction between the technical necessity of a strong grid, having low impedance, and the technical feasibility of dimensioning such a grid to cope with the potential fault current, by allowing a substantial increase in the short-circuit power – without the need to change the existing grid components.
SCFCLs could be installed at different positions in a distribution grid, eg: coupling of busbars via SCFCL; and the use of low impedance transformers in series with SCFCL. By incorporating SCFCL, a customer more than 1 km away from a substation will still be able to benefit from a higher short-circuit power than they could obtain by directly connecting to the busbar. The economical benefit of an SCFCL will depend on the individual situation. But its cost must compete with that of a special line or cable, including switchgear, from the substation to the voltage disturbing customer. But in any case, an SCFCL will represent an economical solution if the need to connect to a higher voltage can be avoided.
SCFCLs are also suited to grid coupling applications on medium and high voltage level, where they eliminate the problem created by the doubling of the potential fault currents, which would otherwise cause enormous upgrading costs.
In addition to power quality and grid coupling, other potential SCFCL applications will include: connection of power stations, especially distributed generation and wind turbines; protection of auxiliary distribution; and protection of older grids which were not designed to cope with high short circuit currents, where the SCFCL may help to postpone a replacement or upgrade.
The next steps
The development of the 6.4 MVA superconducting fault current limiter means that ABB is now close to the 10 MVA level, for which it predicts a substantial demand, which of course depends on the cost of the device. In order to realise an economically viable device ABB will need to further improve its superconducting components and optimise the cooling system of the SCFCL, which is one of the major cost factors.
Thus, commercial availability to electricity operators is still several years away. However, ABB’s long term objective is to achieve the 100 MVA and above level required for the grid coupling applications where the SCFCL technology will really come into its own.