The rapid installation of several large scale solar farms in a remote corner of the NEM has seen yet another example of the challenges this rapid energy transition produces.
AEMO has determined that system strength, a key requirement for a secure grid is so low at times in North Queensland that up to nine solar farms may have to shut down completely if certain other combinations of generators aren’t running. The local transmission operator Powerlink has until 31 August next year to find a solution.
System strength is a technical subject and an area in which Australia is on the “bleeding edge” of international experience. As a report for ARENA explains, our understanding of system strength is evolving along with the changing power system. It’s hard to explain system strength without getting into power system engineering terminology, but here goes. One of the key parameters of electricity is the voltage or electrical force needed to maintain supply (not quite the same as electric power which is measured in watts). In our homes, electricity has a force of around 230 volts and our appliances are tuned to run on that voltage level. As you go back up the power system the voltage gets progressively higher as more electricity needs to be transported. At the bulk transmission level, the voltage is up to around 275,000 volts (275kv). The voltage of an electric system is represented as a wave. This wave needs to be stable for the grid to function. Disturbances, such as changes in the flow of electricity from other parts of the grid, can make the wave temporarily unstable leading to a lower voltage. System strength is the ability of the voltage wave in a given location on the grid to be resilient to nearby disturbances and maintain its stability (if you think all this talk of “force” and “disturbances” is sounding a bit like Star Wars, you’re not alone).
When the grid was powered exclusively by synchronous turbines – spinning machines such as those that power coal, gas and hydro plants – system strength was barely an issue. The rotation of the turbines allows them to “synchronise” with the grid’s waveforms, and the physics of this connection creates allows them to resist a disturbance and limits its size, thereby automatically helping to stabilise network voltages. System strength issues were limited to remote parts of the grid and were well understood by the transmission operators who could apply various fixes to manage the situation.
Solar and wind farms (or big batteries) connect to the grid via an electrical inverter. This uses algorithms to electronically tune the generator to the local wave form, but this doesn’t give the generator the ability to limit a disturbance. In itself this might not matter much, but as these generators displace older synchronous generators, system strength is declining.
More recently, as parts of the grid have seen several wind or solar farms connect nearby over a short period, a new problem has emerged: the interaction of different inverters can make a system weaker and ultimately unstable. It doesn’t help that grid locations that are attractive for renewable generators to connect are often far from synchronous generators and so system strength is already on the low side.
In principle the interaction issue can be resolved by changes to the inverter settings, however as the specifics of the issue are dependent on many variables, such as the make of inverter, the relative location of the generators, the individual settings and so on it can take a lot of investigation by different parties to work out the solution. This approach recently worked in solving an issue in a part of Northwest Victoria dubbed the “rhombus of regret” for the teething problems the solar farms built there have faced.
Other solutions include installing a synchronous condenser – this is basically like a generator except it doesn’t produce any power output. This is what is happening in South Australia, where there is now so little synchronous generation that AEMO has to frequently direct gas generators to switch on until the four condensers are operational.
It’s also how many generators are solving the “do no harm” requirements that have been placed on newly connecting generators. Working out how much “harm” a new plant may do is a complex modelling exercise and the modelling may have to be re-run if the generators changes its inverter specifications or another generator applies to connect nearby. This means the requirements can cause months of delay and significant costs for the developer, so they hate these requirements. And in any case, having lots of small condensers installed is very unlikely to be the most efficient approach.
There are also control systems that can be placed on solar and wind farms, but the owners of these plant don’t much like being curtailed and missing out on revenue, so this isn’t a palatable solution in the long term.
A review is under way to consider the best longer term approach. Options range from mandatory requirements on generators to making transmission operators responsible for maintaining system strength to having AEMO run a market – whether in real-time like the energy spot market or through contracting like the emergency reserve tenders. However, it’s not a simple matter to define how much system strength is required in each part of the grid. So, we’ll be dealing in short-term fixes such as the constraints on the Queensland solar farms for a while to come.