Tesla may well have secured a South Australian state government contract for the world’s biggest battery, but it has expressed deep concerns about the long-term future of energy storage technology in the state, based on the draft proposals in the planned energy security target.
South Australia is considered to be, particularly by its own government, a world leader in the adoption of variable renewables. That level is already above 50 per cent, which means it is at the frontline of how such high penetration is absorbed within a grid. The potential for battery and other storage is huge.
Tesla, however, says that the state risks its losing position as a leader on the issue because of the way the state’s EST is designed, and describes its support for “real inertia” over “synthetic inertia” for meeting some essential grid services is akin to supporting paper records over digital files.
“Tesla is very concerned that the current draft Regulations propose long-term barriers on innovation by excluding rapidly evolving fast response solutions such as battery energy storage,” it says in its submission (made one month before it was announced it had won the big battery contract).
The debate over many aspects of the future grid comes down to the definition and capabilities of inertia. Can renewables, producing “synthetic inertia” deliver the same services as gas and coal and other turbines with “real inertia or “kinetic inertia”?
It is a question that has been the subject of fierce debate in the industry and within the ranks of the Australian Energy Market Operator. And it is a rapidly evolving space, as the industry understands better the capabilities of new technologies.
Tesla’s submission is significant, because it addresses that very issue in much detail, and because of its status as the developer of the world’s first purpose-built “gigafactory”, and its status as a leader in electric vehicles, and solar technologies.
Tesla says the EST in South Australia is being designed on the basis that only kinetic inertia – from spinning turbines such as those in gas plants – can provide frequency response.
“Pushing for kinetic inertia over synthetic inertia in the context of the modern electricity market is akin to supporting paper records over digital file,” it says.
“Only the form of the service differs, and it is a relic of traditional electricity generation technologies – not the innovative renewable energy market that South Australia has developed.”
The key issue that South Australia is looking to address through this scheme is maintaining the frequency of the grid. To that extent it does not matter whether the response comes from the spinning mass of a generator or the controls of an inverter – provided the response is quick enough.
“In the context of South Australian grid frequency issues, synthetic inertia is in many ways superior to kinetic inertia. The notion of synthetic inertia generally appears to be misunderstood, so below we aim to address the benefits of inverter based solutions – both generally and specifically in South Australia,” the Tesla submission said.
Indeed, Tesla then delivered a lengthy overview of synthetic inertia, and its potential role in South Australia.
We reprint this in full because we consider it is critical to the future energy debate in Australia, whether it is the EST considered for South Australia, other state-based initiatives, or the proposed federal clean energy target and generator reliability obligations.
As properly highlighted in the AEMC’s Directions Paper1, frequency control of a grid requires both Inertial Response and Fast Frequency Response (FFR) from grid resources. The Inertial Response of a resource is based on the Rate of Change of Frequency (RoCoF), whereas the FFR is based on the changes in grid Frequency. Tesla battery energy storage systems utilize two separate control schemes to provide both FFR and Inertial Response by reacting to the grid frequency changes and its RoCoF, respectively.
Introducing FFR as a technical solution – which the AEMC is currently proposing to do, will allow greater flexibility in the level of RoCoF that can be permitted and, as noted by the AEMC, is a ‘long-term solution to managing frequency in a low-inertia market’. Tesla battery energy storage systems have been successfully deployed in other parts of the world to provide FFR and help with frequency control of the grid. Battery energy storage will be able to provide this service to enable further innovation and improvement in the renewable energy space in the SA grid.
Tesla utilizes a virtual machine operating mode to provide Inertial Response. From the power system dynamics perspective, there is no difference between Spinning Inertial Response of traditional generators and Synthetic Inertial Response of Tesla Powerpacks. Both forms of inertia respond to RoCoF of a grid in a very short timeframe (milliseconds) before other frequency control mechanisms in the grid start to respond.
Right after a frequency event on the grid, traditional generators momentarily absorb or inject energy at a rate proportional to their inertial constant and the RoCoF. While this phenomenon is referred to as “inertia” due to the momentum of the rotating generator mass that provides the source of energy, it is rather the resulting injection or absorption of energy that helps reduce the grid frequency deviations. Therefore, while this inertial response occurs due to the physics of traditional rotating generators, it can be provided by other very-fast responding resources through controlling their output energy based on the RoCoF of the grid.
A key differentiator between traditional spinning and synthetic inertial response is that the characteristics of the inertial response from battery energy storage can be modified in its control software to achieve the optimal frequency response of the grid. Whereas traditional generators spinning inertial responses are constrained by the physical characteristics and design of generators and cannot be modified to achieve the desired grid response.
Battery energy storage also includes bi-directional inverter capabilities, capable of fast injection or absorption of energy from the grid in order to manage the impact of contingency frequency events. Peaking gas capacity does not have this same fast-response bi-directional capability, and as such cannot quickly absorb energy in the event of high frequency events, or during periods of excess generation.
We note that the emergency frequency control schemes for excess generation events was raised by the Minister as a key issue in the South Australian System Security rule change request. Traditional generators cannot provide this fast bi-directional service – and will not be able to provide improved grid services during high generation, high frequency events.
A further example of the superiority of synthetic inertia can be shown when the ratio of renewables to total generation of the grid changes over the course of the day. At higher renewable penetrations, this change in the ratio can cause a dramatic change in the grid’s total inertia (assuming that the renewables don’t provide inertial response as it is the case today) and the grid’s dynamic behaviour. Battery energy storage has adjustable inertial response which provides grid operators with the ability to achieve the desired dynamic behaviour of the grid at any point in time to overcome this issue.
For regions such as South Australia with such a high penetration of renewable generation, inverter based battery energy storage solutions provides dual benefits of improving dispatchability of the local generation fleet and managing frequency issues in a dynamic fashion
Design of the scheme
As a broader issue, Tesla is concerned with the design of the scheme itself and the approach taken to introduce it.
The primary concern with the current proposed approach and legislative package is that it provides a regulatory solution to what is effectively an engineering problem. Further it locks in one technical solution rather than taking a technology agnostic approach, or detailing the technical problem to be solved and allowing the market to determine the best low-cost, low emissions solution.
Tesla is especially concerned by the fact that such a complicated problem will not be supported by technical guidelines. A core component of the approaches taken by the AEMC and AEMO approach is the establishment of detailed technical guidelines outlining the technical specifications required to solve the problem.
In the context of System Security we support a technology agnostic approach which provides inverter based solutions the opportunity to participate, provided they can meet the technical requirements specified. This ensures competition in the market, and will diversify the asset portfolio providing inertia – providing an additional hedge for ongoing system security risks.
Further, we would suggest that the scheme needs designated review points to assess performance of the scheme in the future.