Proof That Big Battery Plants Work: A Review of the Hornsdale Power Reserve’s Performance
By Aniket Bhor on in Solar Power Industry News
In late 2017, South Australia purchased a gigantic battery storage plant. Built by Tesla and owned by the French company Neoen, the Hornsdale Power Reserve (HPR) was the world’s largest battery storage system at the time. Obviously, a huge group of people thought it was a serious waste of money, while the other group insisted that “time will tell”. And it did!
Within just a year of its operation, the reserve successfully avoided load shedding by maintaining the grid frequency through a rapid response to a power shortfall, while also bringing down the costs of frequency control services. Here’s a closer look at how the battery plant met and exceeded certain performance expectations, and what we can learn from it.
The Hornsdale Power Reserve: A Quick Intro
The HPR is a massive battery system located in South Australia, boasting 194 MWh of storage capacity and 150 MW of power supply capability (129 MWh and 100 MW before expansion). To put this into context, the plant can supply energy to about 150,000 Kiwi homes for about an hour (in theory, at least). It is made by Tesla and houses Samsung battery cells, which means we have some reliable names behind the whole installation.
The HPR is located right beside the Hornsdale wind power farm and shares the same 275 kW network point. The SA government’s contract with Neoen reserves some of its capacity for designated system security services.
Under this contract, the battery provides Fast Frequency Response (FFR) to contingency events and participates in the System Integrity Protection Scheme (SIPS) for preventative protection of the Heywood interconnector (a 275 kV overhead line that connects SA and Victoria). In simpler words, the battery is responsible for pumping power into the grid to regulate frequency variations and avoid blackouts due to component failures or other issues typically caused by extreme weather events.
As for the remaining capacity, Neoen can use it for market participation. This means the battery charges during cheap rate hours and sells power at high demand periods when pricing is higher, and Neoen can make money, which it did handsomely.
Alright! Now to the battery’s performance facts!
How the Hornsdale Power Reserve Performed
It is the 16th of November 2019. At around 1 pm, an unexpected malfunction occurs in the Heywood connector’s communication systems. The connector opens two relays, assuming there is a current differential, even when there is none. Suddenly, the interconnector trips, and South Australia is no longer connected to the National Electricity Market (NEM).
In no time, the local frequency shoots up, and electricity prices, which are dynamic, also start climbing rapidly. The Hornsdale Power Reserve kicks into action, supplying power to the local grid within a second. In just a few minutes, the frequency is back to its normal range, and so are the power prices, saving South Australians an estimated AUD 14 million.
This is just one instance of how the HPR provided “Fast Frequency Response (FFR)” to the grid. According to Neoen, the HPR saved Australians over AUD 150 million in just the first two years. But the HPR’s impact goes beyond just frequency control services.
Aurecon, an engineering, design, and advisory company that has been involved with the HPR since its conception, published detailed reports of the battery’s technical and market impact, and the findings were impressive. We dug into the reports and created a summary of how the plant impacted SA’s grid systems and its users. Let’s dive in.
Fast and Accurate Frequency Control Ancillary Service (FCAS)
Power is made up of multiple elements, one of which is frequency. Technically speaking, the frequency of an alternating current flow is the number of times it reverses polarity. In the Aussie and Kiwi grids, the standard frequency is 50 Hz. Every time the power flowing through an interconnector (in this case, the Heywood interconnector) drastically reduces (or increases), the frequency deviates from this 50 Hz mark.
In such a scenario, there are standby power sources that pump power into the interconnector (or absorb excess power from it) to maintain the frequency. These power sources are called Frequency Control Ancillary Services (FCAS). The HPR is now a major FCAS provider in South Australia. But the most notable aspect of the HPR as an FCAS is its speed and accuracy.
While most traditional power plants need several minutes to start supplying power to regulate the frequency, the HPR does this in a fraction of a second – specifically, around 150 ms. On 25 August 2018, SA experienced a large security event that made the grid frequency plummet, but the HPR acted rapidly and restored the frequency to normal levels within a few seconds. Here is a graph showing what happened:
The green line shows how the frequency dropped from 50 Hz to almost 49 Hz, and how the battery supplied power (yellow line) to bring the frequency up back into the frequency deadband (the acceptable frequency range). Notice how the yellow line is almost a mirror image of the green line, which shows the precision of the battery.
Think of this as driving your car uphill in traffic; you need to accelerate just enough to make the car climb slowly. Too slow and it won’t move, too fast and it will hit the car in front, possibly making an angry guy get out with a baseball bat.
Two more events occurred in the following years, testing the HPR’s capability. The first was an islanding event on 16 Nov. 2019, when an interconnector failure caused the islanding of the South Australia grid, which means it was separated from the rest of the nation. The second was an event on 31 Jan. 2020, when a storm damaged the Heywood interconnector. Below are the graphs showing HPR’s response to both events. Take a bow, dear battery!
As seen in the graph, the frequency drop (blue line) begins at zero seconds, crosses the deadband in the first second, activating Fast Frequency Response (FFR) of the battery. The HPR quickly springs into action (yellow line), pulling up the frequency (green line).
Without the HPR, the frequency would fall below 49 Hz, which initiates load shedding (controlled shutdown of power supply to the customers). But with the HPR, the frequency remains above 49 Hz, avoiding load shedding.
But as we mentioned earlier, it’s not just the speed of the HPR that’s noteworthy; it’s also its accuracy that surprises. Below are two more graphs from Aurecon’s modelling, comparing the accuracy of the HPR with that of a steam turbine acting as an FCAS.
See how the yellow line (generator response) tries hard to match the green line (a signal processed by Aurecon to mimic frequency disruption). In the second graph, the battery response (light green line) traces almost the same line as the artificial signal, meaning there’s virtually no lag or error in the response. This helps maintain the system frequency within the normal 50 +/- 0.15 Hz range.
And before we go to the next section, here’s one last graph that shows the speed of the HPR in responding to a frequency disruption.
The graph shows the time needed for the HPR to start from standby (zero power supply) to a 100 MW supply. And this feat happens in a mere 134 milliseconds, or just over a tenth of a second. This is faster than the average time needed for someone to blink their eyes (about 150 ms). In comparison, most other FCAS providers in the market have a response time of about six seconds.
Alright, now that you’re convinced about the speed and accuracy of battery plants to maintain frequency, let’s move on to the money part.
Cost Savings Through Improved FCAS
The Hornsdale Power Reserve was built for a cost of 56 million Euros, or about $90 million. Naturally, everyone was wondering if the plant would save or make enough money to justify this cost. After all, $90 M is a lot of money that can be used for a lot of useful things. For instance, NZ just promised $100 M to upgrade hospital facilities nationwide.
Fortunately, the HPR made enough of a financial impact to make that price tag worth it. Specifically, it has helped reduce FCAS costs. The Australian Energy Market Operator (AEMO) mandated a local regulation service in South Australia, under which a minimum of 35 megawatts should be reserved for FCAS.
Before HPR, this 35 MW was made up of limited generators, which drove prices up. According to Aurecon’s report, the costs of the required 35 MW FCAS procurement added about 40 million AUD in power expenses in 2016 and 2017 (and about $47 M since its introduction in 2015). When HPR came into effect, it displaced about 10% of the FCAS market, raising competition and reducing pricing. In late 2017, the FCAS costs dropped to a mere $8 M thanks to HPR. Below is a graph showing how the HPR continues to suppress FCAS costs to this day.
But this impact is best understood with a recent islanding event. Let’s take the islanding event of 16 November 2019. Typically, an event like this drives the prices up like they are on steroids, but the HPR avoided a cost increase of about $14 million. Here’s a graph showing the same:
The green portions of the bars show the actual cost owing to the islanding event, while the black portions show the ‘modelled’ costs that would have occurred if the HPR didn’t exist.
Here’s a graphical representation of how HPR consistently pushed down FCAS costs since its operation started:
Notice how the black bars before HPR are tall, which means high costs of FCAS. Post HPR installation, the green bars are much shorter, meaning drastically lower costs. And now our last graph:
It shows the price per MWh of energy supplied as FCAS before and after HPR came online. The green area is pricing before HPR, which goes well above $10,000/MWh. After HPR, this remains at a remarkably low $250/MWh. This reminds me of how one megabyte of storage cost $10,000 in the 1950s and has now dropped below a cent. And now, onto the last known benefit of big batteries!
Improved Renewable Energy Integration
One of the major benefits of the HPR, or any big battery plant for that matter, is its ability to help integrate renewable energy sources in the mix. This benefit has just started showing itself, but in the near future, it may turn out to be the biggest strength of big batteries.
As renewables become cheaper and the planet becomes hotter, the shift to clean energy is inevitable and already underway. A key problem with renewable energy sources is that they are not slaves to our will. Solar panels will generate maximum power when there’s clear sunlight, not when we use more power. The same is true for wind power and most other green sources. And on some days, my toddler’s moods are more predictable than sunlight or wind.
The best way to maximize the use of renewables is to store excess energy when available and use it when the demand exceeds supply. But, and this is a big ‘but’, storing sunlight or wind is not as easy as storing coal. This is where big battery shines – although it cannot store sunlight or wind, it can convert the electrical energy generated into chemical energy and store it in the battery cells for as long as you like.
As more and more big batteries pop up, the use of solar and wind in the energy mix will increase, saving both our big blue home and our money. The same thing applies to electric vehicles. Without batteries, a world full of EVs poses a bit of a challenge for the grid.
Imagine everyone coming home in the evenings and plugging in their EVs for charging. There’s no sunlight for solar power generation, and power demand peaks, leading to several possible problems ranging from frequency deviation to price increase. Batteries can store surplus power during the day and charge EVs at night without posing a risk to the grid’s components or our wallets.
Conclusion
On 18 September 2016, a severe once-in-50-years storm wreaked havoc on South Australia’s electric grid, causing multiple transmission towers to collapse. These damaged towers led to voltage disturbances, causing several wind farms to automatically trip because of their safety mechanisms.
This caused a sudden influx of power from Victoria to SA through the Heywood interconnector. The flow exceeded the interconnector’s capacity, tripping it off and causing a blackout in the state.
The blackout lasted several hours, and it took two weeks to restore power to all areas of the state. The blackout cost the state’s businesses $367 million. Analysis shows that the HPR can prevent, or at least minimize the impact of an event like this. In part, this event was what led the state to get the battery built on such short notice.
And the HPR has since proved its worth. It has successfully controlled frequency variations, significantly reduced FCAS costs, and prevented the tripping of the interconnector as well as subsequent load sheddings/blackouts. While doing this, it has also made good money for Neoen (though it doesn’t really matter to us).
Overall, the HPR has been a perfect example of a zero-sum game, benefiting all the parties involved, particularly electricity customers in SA. This plant, standing quietly in the desert, is glaring proof that big battery plants work, and that batteries are inevitable in our big shift to clean energy.