Last week's Economist included its latest Technology Quarterly report, this time on a technology topic near and dear to my heart – electricity grids: the ultimate supply chains. The five articles in it are worth reading, and as usual with the Economist's Technology Quarterly they do a good job of explaining technologies for a non-technologist reader, from high voltage direct current transmission, to storage and demand flexibility, to "grid-forming" inverters. All of these technologies are inherently digital, every.single.one.
One of the themes of the report is that we have come up with some pretty neat technologies for electric service that's reliable and cleaner; another is that systemic change is difficult to accomplish and takes time and expense, but is ultimately worth it. Understanding these technologies better is a valuable input into better economic analysis and better policy.
I love how the lead article on grid transformation describes the complexity of electrical grids and what's required to deliver safe, reliable, and affordable power to end users:
Within Drax [a large British power plant], energy flows from fuel to flame to steam to spin; but to serve the world beyond the walls a last transformation is needed. The turbines’ driveshafts spin powerful magnets which are encircled by the copper wires of an electric generator. As the spinning magnets’ poles point first this way then that, their magnetic field pushes and pulls at the electrons in the generator’s wires, setting them aquiver with energy. This electromagnetic coupling bleeds off the turbine’s kinetic energy at exactly the same rate as the high-pressure steam replenishes it, thus making that energy available to anything connected to the generator in an electric circuit.
For Drax, that circuit is Britain’s national grid. The 50 cycles a second (50hz) alternating current (ac) power available from pretty much every socket in the country is a national expression of the vibrating floors in Drax and other powerhouses. Its pulsations unite the spinning generators which feed it and all the devices plugged into it into a single vast machine.
This physical process has been the foundation of electricity grids as the ultimate supply chains for over a century – large-scale wires networks connecting large-scale generators and diverse, heterogeneous end user premises like houses, apartment buildings, office buildings, stores, and factories. And connecting them in a network that requires real-time supply-demand balance to keep it from crashing. This daunting and complex challenge is why electricity grids are such a marvel, making them the ultimate supply chains.
Source: The Economist Technology Quarterly (print edition)
Economic, technological, and policy changes are calling into question some of the most important legacy assumptions embedded how we build, operate, and pay for grids. Digitization has changed how, and how much, we consume electricity, and as grids digitize that will change how we produce electricity too, and how we manage the grid to keep it safe and reliable. Economic, technological, and policy forces are shaping the changing generation portfolio, adding low-carbon and resilient to the 20th century objectives of safe, reliable, and affordable power.
At present, 62% of the energy delivered as electricity comes from fossil fuels; that has to come down to more or less zero. A lot of its replacement will be in the form of cheap wind and solar, and that presents a serious challenge to grid operators. It means a lot of new connections, which are troublesome. The problem is exacerbated by the fact that renewable installations typically generate less power than steam turbines do. That means more connections per unit of capacity.
As well as adding a great many new connections, grids will also have to change shape. The places best suited to the generation of renewable energy in very large amounts are often not the places where today’s generation is concentrated. So new transmission lines will be needed. And because grids are complicated things, some of those expansions will require compensating changes elsewhere as bits of the grid become congested. ...
Last of all, various ways in which grids are controlled and balanced today are physically rooted in the way steam turbines generate power. They will need to be rethought. In the long run this is a welcome opportunity to make the system cheaper and more reliable. In the short term it is a requirement for yet more investment.
The last essay in the report, Back in black: The physics of rotating masses can no longer define the electric grid, expands on this idea of redesigning what the grid does and how it does it. The essay digs in to an admittedly arcane aspect of power systems engineering: the motion of the rotation of large generators is the foundation of the grid's frequency and voltage stability.
The synchronisation between the spinning steam turbines of coal, gas, hydro and nuclear plants and the grid they supply is a two-way street: the electromagnetic fields which couple them mean that conditions on the grid reach into the workings of the generators, and vice versa. This means properties of the spinning metal and its connections propagate out onto the grid. One such property is inertia; the turbines’ innate desire to keep spinning limits the ease with which the grid’s frequency can fluctuate. Another is “reactive power”, a drag which the nature of alternating current imposes on the flow of energy through the system, and “short-circuit current”. Reactive power can be used to deal with voltage fluctuations. Short-circuit currents reveal faults and can be used to clear them. Because these aspects of the grid-as-it-is are so useful to its operation, they are referred to as ancillary services.
A big upside of the evolution toward more renewable resources plus storage is the reduction in the need to have generators spinning, "spinning reserves", to make sure the ancillary services are good. These resources, especially storage, could provide ancillary services. If the storage component isn't there all bets are off, since wind and solar are intermittent and need to have backup (usually natural gas generation) for reliability purposes, and those gas generators can provide spinning reserves too, but it's economically and environmentally wasteful to keep a gas turbine spinning in the event that you might need some power from it. A big downside of more renewable resources is the loss of the conventional ancillary services of inertia, reactive power, and short-circuit current that have kept the grid stable in the design we've used for over the past century.
What if we can assure grid stability some other way? We can, through more electronics that substitute for the old effects of machinery. Wind, solar, and batteries are direct current (DC) devices interconnecting on the AC grid through inverters. Digital innovations like grid-forming inverters change how resources "shake hands" with the AC grid:
This means that they can be programmed to provide the grid with energy in exactly the form and at the frequency that the grid operators require, making up for the loss of ancillary services. Grid-forming inverters offer a step change away from the world of instantiated electromagnetism and into a realm of code and electronics.
The hardware which runs grid-forming systems is, for the most part, little different from that in grid-following systems—but the algorithms which shape the current that flows through them are much more sophisticated. And the approach does not have to be limited to batteries. In time all the inverters in front of wind farms and solar plants could all be grid-forming; in some cases, according to Mr Koproski, the change could require nothing more than a software update. In terms of grid stability, this would turn IRBs [sic, it's IBRs] from a problem into a solution. Mr Koproski sees this as turning an old saw about renewables on its head. With the right electronics, adding renewables and the storage which comes along with them to the grid can make it more stable, not less.
Electronics and automation can make the grid easier and cheaper to operate by having the inverter algorithms do what is required in that location to maintain grid stability. It will require changing grid architecture, changing investments, changing regulatory treatment of those investments. The Economist author is hopeful that it will yield a cleaner and cheaper grid:
... the world’s grids have to change if the world is to decarbonise at the rate climate policies demand. That change will necessarily be complex and costly, whatever the technology, with investment measured in tens of trillions of dollars. But it is worth noting that, done properly, this huge and necessary shift will not simply allow the world to continue as it did when burning fossil fuels. By making energy easier to move around than ever before and allowing the most cost-efficient generation to capture more of the market, it will over time make that power cheaper. Robust grids to which cheap generation can be added easily will be able to provide an energy abundance today’s fuels never could.
This set of essays opens up as many questions as it resolves: Does it matter than traditional generation is not intermittent and is therefore more reliable [ed.: is that true?]? How much do all of these system changes cost, and are they worth the investment? If so, who pays, who receives the benefits, and in the case of something like transmission investment, how best to align local community incentives with building that transmission? What legacy assumptions are embedded in "the way we've always done it" that we really should question carefully? Understanding these new technologies and their capabilities inform our thinking through those questions.
Thank you for this explainer. I think most don't appreciate the complexity of the infrastructure that supplies their home or business.
Hmm. Now I understand why State Grid's AI lab is in the world's Top Ten. I'm so glad we sold them our UHV transmission technology when we did. Saved a fortune on R&D. We win again!