Imagine a blazingly hot day in central China, when all the air conditioners in every megacity are running at full blast. Through the remote mountains of Shanxi province, the major transmission lines that carry ultrahigh-voltage electricity to the cities are operating at close to maximum capacity. Heated by the sunshine and the flowing current, the transmission lines sag dangerously close to the treetops. Suddenly the current jumps from line to tree branch, finding the path of least resistance and pouring through the tree into the ground. There’s a bright flash as the current ionizes the air.
During this short circuit, the abruptly unleashed current reaches 10 to 20 times its normal level within a blink of an eye. Now the power grid’s protection system must act fast. Within milliseconds, protection relays must recognize the fault and command the circuit breakers at both ends of the line to switch off the current, isolating the faulted line. The stakes are high: A sustained short-circuit current can trigger a chain reaction of failures throughout the grid and cause widespread blackouts, severely damaging expensive equipment in the process. The 2003 blackout in northeast North America was set off by a tree’s contact with transmission lines in Ohio, which caused a cascade of failures that shut down more than 260 power plants, stopped the flow of 60,000 megawatts throughout the northeast grid, and darkened New York City.
In our hypothetical Chinese short circuit, everything is riding on the action of huge circuit breakers. Just like a household circuit breaker, these industrial breakers open their contacts within a fraction of a second, but because of the enormous amount of energy in the system, just separating the contacts doesn’t stop the current. Instead, the current creates an electrical arc inside the breaker. That small space, which has a volume of just a few liters, now contains a roiling plasma that may reach temperatures of many thousands of degrees Celsius. The breaker can’t contain that plasma for long; if it’s not cleared away quickly, there will be a terrible explosion.
Now the alternating nature of the AC current comes into play: Each time it changes direction (every 10 milliseconds in China’s 50-hertz system), the current temporarily becomes zero, and the energy supply to the arc plasma momentarily halts. It’s at one of these “current zero” moments that the fault current must be interrupted. At that crucial moment, a cooling system inside the circuit breaker injects a high-pressure jet of gas into the gap, blasting away any residue of the hot arc plasma.
Immediately after the arc disappears and the fault is cleared, the power system ramps up again. In this recovery process, the voltage across the gap steeply rises to over 1 million volts before settling to its normal operational level. So in the microseconds before and after current zero, the contacts need to change over from channeling approximately 50 kiloamperes of current through the arc plasma to withstanding 1 megavolt of voltage. This rapid change puts enormous strain on the breakers’ components.
Yet the circuit breakers must perform flawlessly, because the transmission line needs to go back into operation. They must work even though they may have been inactive for long stretches of time and through all kinds of weather. So how can the grid operator in our Chinese example trust that these breakers will do their jobs, and ensure that a megacity won’t end up in the dark? Only rigorous testing can provide that peace of mind. I serve as innovation director for KEMA Laboratories, a Dutch division of the Norwegian consulting and certification company DNV GL. Our task: mimicking the stressful operating conditions of an ultrahigh-voltage AC system in extremis. Replicating this environment is a tremendously difficult engineering challenge, but it’s one that must be met if we’re to satisfy the energy demands of the coming decades.
Tomorrow’s power grid will likely rely on large-scale renewable energy facilities such as hydropower plants, solar parks, and offshore wind farms, located far from power-hungry cities. To transport that energy across long distances, system operators are planning and constructing massive transmission lines. These lines must be high voltage, so they’ll lose only a small fraction of energy through resistance in the lines. Building these cutting-edge high-voltage systems is quite expensive. But many power companies are deciding that the ability to move huge amounts of energy across vast distances justifies the costs.
Choosing to construct a high-voltage transmission system is the first step. The next step is to decide: DC or AC? High-voltage DC transmission systems are an increasingly attractive option, as DC overhead transmission lines require less space and lose less power than do AC lines. But AC technology is more mature, and the world’s most powerful transmission systems are still designed for AC. The latest AC supergrids use ultrahigh voltage (UHV) of at least 1,000 kilovolts, a staggering level not yet realized in DC. In this article, I’ll focus on the gear required for AC networks.
The first commercial UHV-AC grid segment went into operation in China in January 2009. The State Grid Corp. of China spent 5.7 billion Chinese yuan (about US $900 million) on this 1,100-kV project, a 640-kilometer overhead line that connects China’s north and central power grids. A total of 1,284 towers, each about 10 times as tall as the Great Wall, rise above the landscape of the Chinese interior and send electricity across both the Yellow and Han rivers. The towers support 25,000 metric tons of steel-reinforced aluminum conductors that can transport 5,000 MW. The system’s three substations contain circuit breakers capable of interrupting a short-circuit current of up to 63 kA. In 2013, State Grid commissioned a similarly impressive east-west UHV-AC line stretching 650 km between Huainan and Shanghai, which will bring power from inland coal plants to the coastal cities.
Meanwhile, India is seeking to set a new voltage record by constructing a 1,200-kV AC supergrid. In 2012, India’s Power Grid Corp. commissioned a test station for its UHV equipment, and a 350-km section of the 400-kV Wardha-Aurangabad line is now being upgraded to 1,200 kV. The country’s power supply imbalances make such UHV projects particularly beneficial. The Wardha-Aurangabad line will bring electricity from coal-fired power plants in the center of the country to the city of Aurangabad, an emerging IT and manufacturing hub that struggles with power shortages.
We will also see pioneering projects in the decades to come that combine high-voltage AC and DC transmission networks into a “hybrid grid.” (IEEE Spectrum explored this possibility in its August 2015 article “A Globe-Spanning Supergrid.”) In Europe and China, and to a lesser extent in the United States, there’s talk of hybrid systems that would use the advantages of both approaches. For example, a high-voltage undersea DC cable could efficiently take energy from a remote North Sea wind farm and bring it to an AC supergrid, where that energy could easily be transformed and shunted around Europe as needed.
In UHV transmission systems, the most crucial piece of technology is the circuit breaker. The breaker is the system’s guardian: It must be eternally vigilant and prepared to act instantly. And it must function in all environmental conditions and despite great systemwide stress. At the KEMA test facility, in Arnhem, Netherlands, we put these breakers under extraordinary strain to provide an independent assessment of their performance. There’s a clear need for this service: About a quarter of the circuit breakers brought to our labs fail to pass their tests.
Why not rely on simulations to study the stresses at work? Unfortunately, computer models aren’t yet up to the task of simulating microsecond-scale interactions between electrical circuits and extremely hot and chemically complex plasmas. A study carried out by CIGRÉ, the International Council on Large Electric Systems, evaluated the simulation tools used by seven major manufacturers. First, the good news: These different tools did model the electrical fields at critical locations inside a circuit breaker with great accuracy and agreement. But when the tools modeled a breaker’s failure—the point at which it succumbed to electric stress—they produced values quite different from each other and from the true tested value. It’s like modeling the bending of a toothpick: It’s easy to calculate the internal stresses, but the moment and the location of the wood’s fracture can’t be precisely predicted.
Therefore, at our lab we create real-world conditions to determine how the breakers will behave in the field. Of course, the electricity coming into our lab isn’t quite what our tests demand, so we had to develop a few clever tricks that allow us to unleash powerful surges of both current and voltage. In a two-step test process, we mimic the two potentially catastrophic electrical stresses on a UHV circuit breaker in a precisely timed sequence.
First, short-circuit current must flow through the breaker while its contacts separate, thereby establishing an electrical arc inside. At present, we generate this current with four generators, each of which has a 54-metric-ton rotor moving at a speed matching the desired AC frequency, from 16.7 to 60 Hz; very soon another two will be spinning. To initiate the short circuit, 12 synchronized switches energize the circuit, converting mechanical energy stored in the generators’ rotors into electrical energy. We can draw a maximum current of 100 kA from each generator, enough to match the 80- to 90-kA short-circuit currents seen in the world’s most powerful transmission networks.
So now we’ve reproduced the ruinously high currents seen in the first instant of a short circuit. But the breaker’s work is not yet done. It must use its jet of gas to clear the arc from the gap at the critical current-zero moment and then, within microseconds, begin the circuit’s recovery. During the recovery step, the components must withstand a transient voltage surge that far exceeds the circuit’s typical voltage.
The 100-kA short-circuit current that we produce is available at only 17 kV. Half a dozen specialized transformers raise this voltage to 250 kV, but this is still far too low to properly test a UHV-rated circuit breaker. Using additional transformers to step up the voltage further doesn’t make sense because the current would decrease accordingly. So we had to find another way.
For our second trick, we employ capacitor banks as large as four-story buildings, which are precharged to some 700 to 800 kV. At the critical moment, we trigger a spark gap, which discharges the capacitors in sequence. The first bank of capacitors supplies an initial megavolt, and then a few hundred microseconds later, a second capacitor bank adds another megavolt. That’s how we strike the UHV breaker with a voltage that replicates what it would encounter in the field.
In 2008, we tested circuit breakers matching the Chinese supergrid’s specifications using a pilot UHV installation, which supplied 2 MV within the millisecond following a short-circuit interruption. Now we are building a permanent installation, at a cost of $80 million, that will allow us to test both UHV circuit breakers and another key component of the supergrid: high-voltage transformers. Our tests have shown that around 25 percent of these transformers are damaged internally by the tremendous electrodynamic forces associated with short circuits. These transformers must survive the short-circuit current that flows through them for the brief span of time before the breakers do their job, which is no easy task.
It may surprise some electrical engineers to learn that such high-tech test equipment must be brought to bear on something as common as a circuit breaker. Didn’t we figure out everything there is to know about breaker technology decades ago? In fact, the technology is still evolving, and so must our ability to evaluate its performance.
Oil-based circuit breakers dominated in the early 20th century. In these devices, the contacts sit inside a tank filled with oil; when the arc forms, it turns some of the oil into a high-pressure gas bubble, which surrounds and extinguishes the arc. But these oil tanks are unwieldy and dangerous contraptions. The 1970s saw the rise of circuit breakers that use sulfur hexafluoride, an inert gas with good insulating properties, which is blasted through the gap to quench the arc. However, SF6 is an extremely potent greenhouse gas, so the electricity industry is now developing alternative technologies.
Many researchers are investigating a type of circuit breaker that interrupts the current in a vacuum environment. The main difficulty here is managing the electric field in the vacuum. Because no gas or liquid is present, the arc plasma creates its own medium by releasing and ionizing metal vapor from the contacts themselves. As the hot plasma burns the contacts, it deforms their surfaces, creating microscopic dips and spikes. The protrusions jutting from a contact’s surface are analogous to tall trees that tower above Earth’s surface and are thus more likely to be struck by lightning. The roughened contacts may continue to channel current when they shouldn’t, namely when the breaker is attempting to clear the arc.
To further develop these vacuum-based breakers, we need to test them under full-power conditions and study the current of mere tens of microamperes that is drawn out of the contacts’ protrusions by the quantum mechanical tunneling effect. Do these tiny currents signal an imminent resumption of the electrical arc and thus the failure of the breaker? That’s a heavily debated question in the scientific community. At KEMA Labs, we’re looking for answers by evaluating the effect of these tiny currents in real equipment under full stress.
Indeed, many of our research projects involve working on the smallest scales imaginable. Surprisingly, very fast processes occurring on the microscale often determine whether the massive components of a transmission grid will malfunction—and possibly whether an entire city will go dark.
We must be able to study, for example, the events within the circuit breaker during the few microseconds surrounding the current-zero moment. In that minuscule time span, the breaker must change from a very good conductor to a nearly perfect insulator. With the newest tools, we are now able to monitor this transition. One high-resolution method we’ve developed can detect currents smaller than a single ampere and lasting just a few microseconds during full-scale short-circuit tests, which use currents that are measured in hundreds of kiloamperes. We look for these minute currents after the current-zero moment during which the arc should have been totally extinguished. If we find them, we have an indication that something is wrong with the breaker’s recovery, and that the full electric arc may blaze anew inside its gap.
The electricity supergrids of the 21st century will rely on these circuit-breaking guardians, stationed like sentinels along continent-spanning transmission lines. And the breakers in turn rely on engineers in the test labs who unleash powerful floods of electricity to stress these components to the max, while also probing the intimate processes within the breakers’ hearts. Only such top-to-bottom testing will ensure that the equipment can keep a Chinese megacity cool on even the hottest day.
This article originally appeared in print as “Safeguarding the Supergrid.”
About the Author
René Smeets decided to be an electrician when he was a 10-year-old boy, gawking at the smoke pouring out of a busted circuit breaker panel in his home. Later, after studying physics and obtaining a PhD, he amended his plan slightly, opting instead to research industrial-scale electrical transmission systems. He currently serves as director of innovation at KEMA Laboratories, a division of the consulting and certification company DNV GL, where he oversees the stress-testing of circuit breakers and other components designed for ultrahigh-voltage transmission networks. He is an IEEE fellow and author of a book about switchgear.