How To Solve a Problem Like Grid Stability
Inertia has provided frequency stability since the introduction of synchronous generators in the first AC electricity networks. But as we increase the role of non-synchronous generation in the grid, we face a system-stability challenge.
As the grid becomes distributed, diversified, and less synchronous, we need to design new ways to stabilise it. Stability needs to be distributed and built into the grid itself, in a highly efficient and non-intrusive manner.
Thirty years ago, I went to a remarkable concert in which Bobby McFerrin had 3000 people in the audience sing Ave Maria while McFerrin sang the Bach prelude that Gounod used as a basis for Ave Maria. The experience is one of my strongest musical memories. The vocal skills of the 3000 audience members obviously varied considerably.
While the tune of Ave Maria, the “signal”, came through clearly, there was as softness around the edges as people sang off tempo and out of key, and occasional mistakes popped through as transient noises and variations in pace. It was a beautiful and unique experience, but I wouldn’t want every concert I attend to be sung this way.
I’ve written before that understanding electricity as a wave, not a commodity, is key to unlocking a successful energy transition. The (Nikola) Tesla-Westinghouse AC electricity system that has powered the last century of growth was like listening to an opera: a small number of powerful, highly trained voices producing a clear, harmonious signal.
By contrast, the renewable dominated system that we now have, with more than a million points of generation in the UK alone, is more like McFerrin’s Ave Maria. Solar and wind farms are not synchronously coupled to the AC waveform in the grid. Rather, they use inverters, essentially switches, to create a simulacrum of the AC waveform and inject that into the network. They are the like the audience members trying imperfectly to keep up with the tune. There is a clear signal, in part generated by powerful thermal generators that play the role that McFerrin himself does in Ave Maria, but that signal is surrounded by noise, harmonics and transients. Anyone with a good hi-fi can literally hear this noise in the electricity system as static when they turn on their receiver. Importantly, just as the singers who are out of key are still producing sound, the noise is energy, just energy in the wrong place in the AC waveform.
Sadly, noise, harmonics and transients, created in the system create much bigger problems than just static in your stereo. They reduce the carrying capacity of the electricity system. Some estimate that fully correcting the signal would meaningfully increase network capacity with existing poles and wires[1][2].
These system harmonics and transients also damage equipment throughout the network and create power quality problems. Data centres regularly struggle with power quality, and transformers that would previously have lasted 30-40 years, now often fail after only 20. Small turbines suffer bearing and blade failures, as do many other types of sensitive machinery. We need a solution to the problem of noise.
In Ave Maria, if the audience goes off tempo, McFerrin’s pace brings them back in. Inertia in the electricity system plays a similar role in stabilising the frequency of the system. As the level of inertia falls, frequency—the tempo of the AC waveform in the grid—is able to deviate further, faster from the target frequency of 50hz or 60hz. Given that the frequency must be kept within +/- 1% of the target at all times, this increase in the rate of change of frequency (ROCOF) caused by lower levels of inertia, is a serious problem.
Electrical grids need to be kept stable at all times to ensure that when you flick a switch, the light turns on. Today, that happens by maintaining system frequency, in a simplified description of power/energy balance of the grid. Inertia is a property of synchronous machines, most commonly the carbon-intensive thermal generators of power stations. These machines are said to be “synchronous” because they are magnetically coupled to the waveform in the grid, and their speed of rotation is directly linked to the frequency of the electrical signal they generate, i.e. the 50hz target frequency of the European grid matches the turbine of a generator spinning at 3000rpm. If the machines accelerate, the frequency increases, and vice-versa. The converse is also true, i.e. if the system forces the frequency to increase, all synchronous machines that sense this frequency will accelerate, and vice-versa. The synchronous machines dance to the beat of the system and the beat of the system is formed by the behaviour of all the synchronous machines.
Because of its importance, systems with high level of demand or supply volatility maintain what is known as “spinning reserve”. This is done by paying some of the most responsive power stations to stay spinning even when the power from that station is not required.
As the system gets more synchronous, other phenomenon such as noise and harmonics take prominence. The noise and harmonics in the network dissipate as heat, ageing or damaging equipment. It also means that the energy in that noise is lost. We need a new solution for maintaining system stability, that can capture this noise and put it to productive use. This new solution also needs to reduce our dependency on using carbon-intensive power stations to maintain stability.
This is one of the key challenges that needs to be solved, if we are to have a future system that is decarbonised, secure and affordable.
So what are the key strategies available for solving these problems?
1. Keeping renewables out of the system – which simply isn’t a viable solution for a world moving towards Net Zero;
2. Maintaining inertia as an add-on to a renewable-based system – which again, isn’t a viable option if we want to decarbonise our energy system and deliver energy system affordability;
3. Actively correcting the signal to maintain frequency, removing the need for inertia, and putting the electromotive force of noise and harmonics to useful work, as we do at ENODA with our Enoda PRIME® Exchanger.
Keeping noise and inertia-free generation out of the system may seem like an obvious solution, but it means big climate sacrifices. Renewable energy curtailments are already common globally, as seen recently in Scotland, where the nation was plunged into political uncertainty as a result. Variable renewable energy generators are regularly fined in Europe for the pollution they inject into the system. California has repealed its solar roofs initiative. Western Australia has renewable plants lying fallow, and South Africa has ceased connecting new solar farms in the Northern Cape despite regularly suffering swingeing power cuts.
Not only does keeping noise out of the system mean curtailing renewable generation, but it also isn’t possible because the AC waveform is a product of both generation and consumption. Noise, harmonics, and phase imbalances are introduced by consumption as well as by generation. The Tesla-Westinghouse system was designed for a world in which electrical devices were all AC and demand was essentially synchronised. Everyone cooked dinner and watched TV at the same time, and generators could ramp as a major football match reached half time, ready for everyone to open their fridges to get a beer or boil their kettles to make tea before sitting down to watch the second half with a beverage.
Today, the situation is fundamentally different. Many of our electrical devices, like computers and EVs, are DC, running off of converters. Our behaviour is no longer synchronised, and our attempts to address climate change have made it much less synchronised. Not only do neighbours watch television and cook dinner at the same time; one house’s solar panels will come on, reducing its demand, while another home has an EV that pulls as much power as the home itself in the evening when it charges. If those homes are on different phases of the signal, the local transformer that splits the three-phase power into single phase power to individual homes, will experience serious phase imbalances and suffer damage.
We need to accept that noise, harmonics, phase imbalances and transients will be introduced into any modern electricity system and that they will rise as the share of renewables in the network rises and electrifications of load continues. This is why it was so important for us to develop a technology that could deal with the problem of noise – it’s essential if we want to address climate change.
Enoda PRIME® Exchangers can enable the system to operate with very low levels of inertia. Prime Exchangers can correct the AC waveform in real time, enabling them to restore frequency almost instantaneously, closing the system energy gap of any frequency excursion even when ROCOF is high. Prime Exchangers can also correct noise and harmonics. Enoda PRIME® Exchangers can recover the energy that is lost as noise and harmonics and reinject that energy in the correct place in the AC waveform, enabling it to be put to beneficial use.
Other technological solutions simply aren’t viable. Given that noise accounts for 7-9%[3] of the total electromotive force in a network with a high level of renewable energy, dissipating this noise as loss seems, at the very least, incredibly wasteful. To put this in perspective, it is equivalent to the UK discarding half of the nuclear power it generates. The spinning mass of a thermal power station physically resists the change in frequency in the network. Thermal generators do this as they spin, and inertia can be artificially added to a system without them through synchronous condensers, an expensive solution that spins a massive electric motor that is coupled to the AC waveform, so that the mass of the motor can spin and resist frequency change in the network—it’s essentially a thermal generator without any generation—an incredibly inefficient solution to the problem, but one that is currently being rolled out globally.
The amount of noise and harmonics in the network will increase with the share of renewables. The International Energy Agency includes noise and harmonics as part of lost energy. Figure one shows the increase in lost energy from 40% of total generation today to 50% by 2050 as the share of renewables rises form 40% today to 85% in 2050. In actual fact, the compounding effect of noise in the network and disparities between renewable generation patterns and demand mean that losses are likely to be higher without a new approach to the problem. The interaction of harmonics with the physical infrastructure, especially transformers, means that increasing the harmonic load induces an increasing rate of harmonics being created in the grid as the dirty signal interacts with the current infrastructure—a vicious cycle.
Figure 1: IEA forecasts show that the share of electricity generation that is lost will rise as renewables scale up
ENODA’s technology can put this excess electromotive force to work rather than dissipate it as thermal loss. This could make a larger contribution to the UK’s power than the delayed and overbudget Hinkley Point C nuclear power station.
As we increase the role of non-synchronous generation, like solar and wind, we have no choice but to implement new technologies to stabilise the grid. Stability must come from the bottom-up to ensure it is both highly efficient and non-intrusive. This is the only way to ensure a renewables dominated grid won’t face catastrophic failure, thus enabling a viable pathway for the energy transition. Remaining wedded to mechanical inertia as the basis for system stability, will only hold back the energy transition when technologies for active signal correction, like ENODA’s, can offer a much more effective mechanism for stability in the future energy system.
References
[1] UK Power Networks "What are electrical losses?
[2] Yokogawa Test & Measurement, “Power Quality and Energy Efficiency for Power Measurements”
[3] Goran Strbac, et al. “Strategies for reducing losses in distribution networks”, Imperial College London and UK Power Networks February 2018
