Recent developments in the National Electricity Market have caused some solar and wind farms to be curtailed due to “voltage oscillation and system strength issues”.
In addition, the market operator has warned of delays to some new connections. Accordingly, it seems apt to spend a few key strokes describing what could be causing the problem.
Oscillations, voltage or otherwise, are a common problem experienced by feedback control systems of which there are several everyday examples.
E.g. the high-pitched whistling that public announcement systems sometimes produce – which is caused by microphones placed near speakers with the volume turned up too high. The sound from the speakers enters the microphone which then becomes amplified, which then goes out the speakers, back through the microphone continuously leading to a cyclic amplification of sound.
Or, Spinning out a car on a sharp turn because of too high a speed and aggressive turns of the steering wheel
Or, the difficulty of a tight rope walker to maintain balance on a swaying rope. The walker tries to maintain balance by swaying in the opposite direction to the rope, however if he or she makes moves which are too fast or in the wrong direction – then overbalance and fall from the rope will be the outcome.
Or, the boom and bust economic cycle of a modern industrial economy. Many economists have attributed this to the various feedback paths within the economy which have been introduced by global industrialisation and the ability of modern manufacturing to outstrip demand.
With generators – be they traditional rotating machines, or inverter connected plant – it is important to control the voltage. To do this you need some sort of feedback control system. However – as with all systems controlled by feedback – the system can oscillate out of control if an unfortunate association of time delays and feedback gains happens to coincide.
All of this is obvious.
The solution is also obvious – simply change the feedback gains and time delays (many of which are set in the control system software) appropriately to remove the possibility of an oscillation occurring.
So for the case of the new solar farms installed in South Western NSW, and North Western Victoria why didn’t this occur?
A clue can be found in the national electricity rules and the way they have been interpreted and applied – specifically how inverter plant is being required to respond to power system disturbances – which in summary is as fast as possible.
There is no good reason why we need inverter connected plant to respond much more quickly to system disturbances than is currently required of traditional generation, it probably just seemed to someone to be a good idea at the time.
After all, if you are given the choice between a high-performance sports car and slower but more practical sedan – most people would initially opt for the sports car – at least until the common-sense portion of their brain kicks in.
This seems to be what’s happened in this case. It is possible for inverter connected plant to respond much more quickly to voltage disturbances than a synchronous machine can – that doesn’t mean it is always a good idea to do so.
Even the slower traditional synchronous plant has voltage control stability issues. It is common practice to install power system stabilizers on generators – they are specifically designed to slow down the voltage control response to prevent the voltage output from generators from oscillating out of control.
Needless to say – traditional plant with or without power system stabilizers cannot respond as fast as modern power electronics – nor does it need to.
But synchronous plant has its own issues, and simply adding more of it in the form of synchronous condensers to a portion of the grid won’t fix the underlying problems. In fact, it can easily make things worse by adding unnecessary complexity and creating another model which can be restricted by a complex web on non-disclosure agreements.
Whenever new technologies are introduced – there seems to be this transition period where various regulators try to pretend things are the same as what came before.
Hence, we used to limit the speed of cars to about 5-10 miles per hour because that is what horse and carts could typically do. Appropriate road signs took a long time to develop, after all they don’t need to be large and simple if you are speeding along at 5 miles an hour – in those days you could stop to read them. We no longer have that luxury.
Going back further in time – there used to be no standard width for vehicles, or even standardised sides of the road to drive on. It seems to me, the situation we currently find ourselves in has many similarities with this earlier period.
So where do we go from here?
Do we continue with these poorly defined rules for connection of plant? Or do we start making rules which actually take due account of the laws of physics and the physical limitations of the plant.
The recent rule changes do not reflect the laws of physics. In fact, if interpreted literally they make the connection of new plant whether traditional synchronous or inverter connected physically impossible.
As I write this, summer is just around the corner and it looks likely that there won’t be enough generation to cover the peak load.
Whilst the shortfall has been mainly due to equipment failures – it hasn’t been helped by poorly drafted and interpreted registration procedures. These have created delays in the connection of new plant just when we need it most.
