In all cases, one particular feature of superconductors is important: When too much current flows within the superconductor, it changes from offering negligible resistance to substantially resisting any further increase in current flow. This behavior’s role differs from FCL design to FCL design. In the resistive FCL and in the shielded core FCL, each described below, this behavior is crucial to limiting the fault current. In the saturated core FCL, also described below, the superconducting material is meant to stay in the superconducting state while a different mechanism limits the current. Indeed, switches remove the superconductor from its circuit when the change to substantial resistance would otherwise occur.
Resistive FCL
One kind of design – a "resistive FCL" – uses the change in resistance to make a device in which resistance increases to partially compensate for the sudden absence of the load that was circumvented by the unwanted, unanticipated short-circuit (or fault). This prevents the fault current from growing to an unmanageable degree. However, to implement this kind of design, several impediments must be surmounted.
First, the current must be passed from a metal (e.g., aluminum) at ambient temperature to a superconductor operating at much lower temperature, and then it must return to metal at ambient temperature. Thus, the device's terminations must be able to handle both a change in temperature and a change in electric field in the same place. Many teams have achieved this for voltages below 15 kV, but this challenge becomes significantly greater when the voltage is much higher, 115 kV or more.
Second, while serving as a resistor to the fault current, the superconductor's temperature rises. (This is to be expected. The fault current's electromagnetic energy dissipates much more quickly than any refrigerator can reject the resulting heat to the environment.) From a utility's point of view, the vital concern is the time required to cool the superconductor back down to its operating temperature after the fault current is limited. Fast cool-downs are more important in some applications than others, but time is always relevant. Some faults are intermittent (e.g., a tree branch touches a power line, the wind pushes it away, it rebounds, and so on); even though the fault appears to have vanished, it may quickly reappear, making necessary the limiter's repeated operation with only brief breaks in between. Or, without knowing whether the fault has ceased, the utility may (re)close its circuit breakers to ascertain whether the fault persists. Some utilities do this three times in quick succession and then lock the breakers open if the fault persists through the third try. Where such procedures are employed, the FCL's ability to quickly return to service is vital. (Of course, where quickly repeated service is not required, the FCL may be allowed more time to recover.) Various ways to avoid slow cool-downs back to operating temperature have been explored. The most straightforward is to use the superconductor only until a switch can divert the fault current into an ambient-temperature impedance. By this means, the superconductor's temperature need not rise so high as it would if the superconductor had to resist the fault current until the circuit breaker opened. This strategy was adopted by a KEPRI-LSIS team, who report promising performance of their 22-kV FCL and have recently installed it in the Icheon substation near Seoul, where this FCL and an HTS cable will be part of KEPCO's grid. The AMSC-Siemens-Nexans team, which recently completed a successful test of their 115-kV resistive FCL (a significant technical accomplishment), also adopted a like strategy, with satisfying results.
Finally, a comment: utilities will be greatly interested in the reliability of the resistive FCL. Because the line current passes through the cold superconductor, a failure there (e.g., refrigeration) would interrupt the routine current flow and so affect routine power delivery. Of course, the desired reliability will depend on the location of the FCL; if it ties together two bus bars (meant to equalize voltage by small current flows), the FCL need not be as reliable as one on a “feeder line,” which is meant always to deliver power. Thus, KEPCO's demonstration, recently begun in its Icheon substation, will receive close attention.
AMSC-Siemens-Nexans Resistive FCL
The AMSC-Siemens-Nexans prototype is the only one among all resistive FCLs that is intended to address high-voltage (HV) transmission lines. At the outset, it must be understood that this team decided to construct a single-phase prototype, not a three-phase device, and that the prototype was tested at PowerTech Labs Inc. in Surrey, BC (Canada), not on a utility grid. Nonetheless, the success of the test suggests that a lack of resources, rather than any technical issue, impedes the construction of a three-phase device and a subsequent grid demonstration on a 138-kV line.
FCL (1) and LN2-tank (2) in the open-air test bay of the PowerTech high-power lab. The terminations (3) are connected via aluminum rods to the lab's bus (4). (5) Potential converters.
This FCL prototype is intended to limit faults on a single-phase, 60-Hz, 115-kV, 0.9-kA line; the equivalent three-phase line would convey 180 MVA. The FCL prototype's standby or insertion reactance is 12x10-3 
and its resistance is 0.150x10-3 . And, during routine operation the FCL's standby cooling load is approximately 16-kW per phase which maintains its 3.4 km of steel-backed REBaCuO tape at 74K. So as to have small inductance, this tape is wound into bifilar coils. The complete current limiter unit comprising the FCL and a parallel reactor is designed to limit a prospective 63- kA fault current to 40 kA. The limited current is immediately redirected into the air core reactor, and a fast switch in line with the FCL interrupts the current flow through the superconductor within 67x10-3 sec, The FCL can recover from limiting one fault current and be ready for the next in 15-20 seconds, while the limited current can continue flowing through the reactor if required by the installed grid protection relays and method of fault location. With regard to installation, the FCL would comprise three air core reactors and three cryostats (one for each phase). The height and outer diameter of each air core reactor would be approximately 0.55 m and 1.5 m, respectively. Each cryostat has a 10-m length and a 2-m outer diameter. Each of the cryostat’s two HV terminations is 2 m high.
Additional information about the test, as well as about the AMSC-Siemens-Nexans FCL will appear in the proceedings of EUCAS 2011. Also see W. Schmidt, B. Gamble, H-P Kaemer, D. Madura, A. Otto, and W. Romanosky, Design and test of current limiting modules using YBCO coated conductors. Supercond. Sci. Technol. 23, 2010, 014024.
Saturated Core FCL
Another design – the "saturated core FCL"– does not require the line current, either nominal or fault, to pass through the superconductor. Thus, this design can be implemented using conventional terminations, which makes it particularly appealing to those who wish to limit high-voltage fault currents without having to master the skills that Nexans acquired by building the terminations for its 115-kV HTS LIPA cable and the terminations for the 115-kV Siemens-Nexans-AMSC FCL. The saturated core FCL operates in a completely different way from the resistive FCL. Instead of initially dissipating the fault current's power in the superconductor, the saturated core FCL immediately produces an electric field to push back on the fault current, thereby reducing its magnitude to one that the circuit breakers can manage. The "push-back" electric field (or "back EMF") results from reversing the magnetization of iron cores — a reversal that is itself driven by the fault current.
Under no-fault conditions (i.e., during normal operation, with the FCL simply standing by), a way must be found to eliminate the hysteresis loss in the iron cores that would otherwise occur with each cycle of the line current. This can be done by enforcing constant magnetization with the help of a direct current circulating within a coil embracing the cores, which coil is connected to its own separate power source in a circuit wholly separate from the grid. The separate circuit is made from superconductor, for two reasons: (a) in the absence of a fault (i.e., almost all the time), the loss in the superconductor's refrigerator is small compared with the resistive loss that would occur were the coil made from copper coils; and (b) the superconductor required to carry the needed DC current is much smaller and lighter than the copper that would otherwise have to be used.
Two impediments to implementing this design should be recognized, if only to note that recent progress appears to have surmounted both. First, the same reversal of the iron's magnetization that limits the fault current puts a sudden voltage across the superconducting coil, which might damage the coil (by causing part of it to become resistive) and would more likely damage the coil's power supply. One way to avoid this is to very quickly remove the vulnerable power supply from the superconducting circuit as soon as the FCL goes into action. A second impediment has been the volume and weight of the needed iron. Forty years ago, when the design was first implemented, the pioneering team used so much iron that others were discouraged from following up. After a long hiatus, three different teams have recently reduced the needed iron to a modest amount. These same three teams — InnoPower (China), Zenergy (Australia and the U.S.) and GridOn (Israel and Australia) — each claim to have solved the problem of protecting the power supply. Now, each of these teams is making progress toward commercial superconducting FCLs. Indeed, InnoPower is about to install a 220-kV FCL on the grid in China. Zenergy has delivered an FCL to the UK and has contracted to deliver another during 2012, as well as having longstanding plans to deliver a 138-kV FCL to AEP, a U.S. utility. And GridOn has contracted to deliver another saturated core FCL (this one not to be superconducting) to the UK's ETI in 2013.
InnoPower's Saturated Core FCLInnoPower's saturated core FCL is the only one of its kind intended to address the needs of very-high-voltage transmission lines. Factory testing has just been successfully completed, and this FCL is to be installed on the grid at the Shigezhuang substation of Tianjin, China, before the end of 2011.
This FCL is intended to limit faults on a three-phase, 50-Hz, 220-kV, 0.8-kA, 300-MVA line. The FCL's standby or insertion impedance is 1.85 Designed to limit a prospective 50-kA fault current to 30 kA, this FCL can recover from limiting one fault current and be ready for the next in 0.5 seconds. With regard to installation, the FCL is 8 m long, 8 m wide, and 9 m tall; its total weight is roughly 136 tonnes, of which transformer oil (see below) accounts for roughly 16 tonnes.
This FCL's construction is also of interest. The high-voltage line that wraps around the iron cores is insulated from them (and they, from each other) by transformer oil; this oil is contained in fiberglass tanks to avoid losses from the eddy currents that would be induced in metal tanks. The power supply's fast switch is an insulated gate bipolar transistor, IGBT. The high-temperature superconducting coil, with a mass of 800 kg, is made from approximately 5 km of Bi-2223 tape. The coil is cooled by liquid nitrogen that is delivered by truck to on-site storage (10 m3), when needed. When the FCL is standing by, waiting to limit a fault current, the total current circulating in the HTS coil is in the range of 100–300 A (the actual value is chosen by the operator), and the voltage difference across the coil's leads is a fraction of a volt.
Additional information about InnoPower's FCL will appear in the proceedings of EUCAS 2011.
Shielded Core
Another class of designs also separates the superconductor from the nominal line current. Only an inductive coupling is permitted between the grid's fault current and the limiting element (or circuit containing that element). In the absence of a fault, the limiting element is isolated from effects of the line current by a superconducting shield: In this design, the alternating magnetic field of the alternating line current induces eddy currents in the superconducting shield, and the magnetic field of these eddy currents is sufficient to cancel the field of the line currents in the limiting element so that it remains oblivious to the normal operation of line. However, when a fault floods the line with current, its field is so large and changes so quickly that the superconducting shield's eddy currents cannot cancel the fault currents' field. Instead, it reaches into the limiting element, where it induces currents that encounter impedance or in the most straightforward case (see below) it simply sees an impedance; thus, the fault current effectively triggers an impedance by which the fault current is itself limited. As with the saturated core FCL, the shielded core FCL requires only conventional terminations, and its design makes very few demands on the superconductor or its cryogenics when there is no fault (i.e., almost all the time). During the past decade, these features have attracted academic investigation using lab-scale equipment. Now, a Bruker-Schneider Electric-Stadtwerke Augsburg team is making progress toward a commercial superconducting FCL.
Bruker-Schneider Electric-Stadtwerke Augsburg iSFCLIn 2013, the project team plans to install its inductive Superconducting Fault Current Limiter, iSFCL, a first-of-its-kind prototype, between a substation of Stadtwerke Augsburg and a nearby factory, owned by MTU Onsite Energy GmbH, where its products are extensively tested. The iSFCL is intended to limit faults on a three-phase, 10-kV, 0.8-kA line. A prospective 25-kA fault is to be limited to 5 kA on the first peak and soon thereafter to 2 kA. The iSFCL is designed to initiate limitation very quickly, in ?1-ms, and to hold the fault for 0.5 sec.
A one-line diagram of the iSFCL circuit appears below.
The Bruker -Schneider Electric-Stadtwerke Augsburg team is realizing the generic approach described above as follows: a normal metal inductor (the primary coil) is inserted in the line but, instead of having an air core, the primary's core is filled with a hollow superconducting cylinder within which is a solid iron cylinder. The result of this arrangement is that the alternating magnetic field of the nominal alternating line current induces eddy currents in the superconducting shield, and the magnetic field of these eddy currents is sufficient to cancel the magnetic flux in the solid iron cylinder. Thus, in the absence of a fault, the impedance of the device is much less than the primary coil - iron core assembly without the hollow superconducting cylinder. On the other hand, when a fault floods the line with current, the current induced into the superconducting shield exceeds the critical current of the superconductor and so it can no longer shield the interior solid iron core from the fault current's magnetic field. In the absence of shielding, the fault current encounters the impedance of the primary coil and the iron core. In this way, the fault current triggers a suddenly exposed impedance by which the fault current is itself limited.
The relative size of the components are illustrated in following drawing . The simplicity of the conception is notable and so the forthcoming demonstration deserves particular attention.
The superconducting cylinder is fabricated from REBaCuO tapes. A fiberglass cryostat is used to avoid the loss that would otherwise occur in a metal cryostat. One such device will be put on each of the three phases of the line connecting of Stadtwerke Augsburg and the nearby factory, owned by MTU Onsite Energy GmbH
Additional information about Bruker's iFCL was presented in talk 2-LAO4 at EUCAS 2011.
