Chris Morris is a semi-retired power station engineer in New Zealand who has commented here on No Minister occasionally and on other NZ blogs. In mid 2023 he emailed me about a series of four articles he had written for the blog of Judith Curry in Australia.

I published a summary of the key points of his first article here, Australia’s Transition to Renewable Energy, Part 1, which was an overview of the Aussie system and Part 2, which looked specifically at South Australia(SA). Chris’s Part 3 deals with the impact of renewable electricity on the grid and follows on from the point made in part 2 that replacing fossil fuel generators with renewables is easier than building a grid to cope with them.

There’s some “esoteric” Power Grid 101 stuff at the start that laypeople can ignore aside from six key things:

  1. Active Power is what does work (measured in Watts).
  2. Reactive Power is the energy in electromagnetic fields in and around electrical equipment. It’s measured in VARs (Volt Amps Reactance).
  3. Both operate at cycles of 50Hz in Australia (a Hertz is one cycle per second). The grid doesn’t perfectly sit at 50Hz all the time, it wobbles a bit and if it wobbles too far below 50Hz the system collapses.
  4. The two have to be balanced for the whole grid to work (and not collapse).
  5. Transmission lines both consume and produce Reactive Power. When lines are lightly loaded, they are like capacitors (negative VARs) so absorb VARs, at the SIL rating (Surge Impedance Load, see graph below) they’re neutral, and at heavy loads they’re like inductors (positive VARs)
  6. Electric motors also produce a lot of Reactive Power that the grid has to absorb.

As a result the balancing act gets tricky. The huge 2003 blackout in USA/ Canada was caused because they weren’t managing Reactive Power properly.

Wind and Solar power makes all this much more complicated – hydro and geothermal are good baseload, synchronous power generators that are renewable but they don’t count for much in Aussie:

  • Long transmission lines are needed because renewables are often sited a long way from where the power is used.
  • The loading is far more unpredictable than with conventional baseload power generators, so Reactive Power goes up and down a lot.
  • This means the switchyards along the line need extra, expensive equipment to manage the changes and stabilise the system.

Any grid has to cope with things like lightning strikes, transmission towers being damaged by high winds or floods, or generating plants that just go offline for any number of reasons. That’s when balancing the grid gets even trickier because fast action is needed to prevent that all-important frequency from dropping too far below 50Hz. To that end there are three crucial factors:

  • Inertia in the system, provided by the rotating machinery of a generator, is important because the more of it you have the slower the rate of change in frequency when something does go wrong.
  • Reserves of power that can be cranked up fast to put into the grid, “dispatched” is the word. The amount of reserve needed is usually sized for the loss of the largest supply item on the grid.
  • Load Shedding – literally cutting off parts of the grid to prevent grid collapse – has to be done if you lack sufficient Inertia and Reserve or if they can’t react fast enough. Load shedding is largely automatic nowadays.

As more renewables have been added to the Australian grid the following has resulted:

  • They throw the frequency by both producing little or no power, but also sometimes too much. There is less Inertia in the system.
  • The Reserve used to be fossil-fuel plants humming away in the background 24/7, but as they’ve been shut down it’s now increasingly supplied by gas-turbine units which can ramp up and down very quickly. But they can’t stop and start instantly either; they have to be kept running constantly at a low level, which is extra cost, and of course that plus their capital cost is having to be repaid from smaller amounts of electricity than if they ran 24/7. Batteries can do the job in theory but are even more expensive.
  • As a result the swings in grid frequency have become more rapid.
  • So the faster the dispatch response time has had to become. The minimum used to be 6 seconds, now a 1 second response is being proposed and that costs money.
    [Frequency control] has dropped back to about $40M a quarter. In 2010, before there was significant wind or solar on the grid, the FCAS cost was about $2M a quarter. [a 2000% increase] FCAS costs are operational charges to be paid by consumers for the privilege of the renewables penetration.

The article looking at how the Queensland grid, with little renewable energy, coped with problems, compared to how South Australia did ith similar problems.

In 2021 a number of Queensland coal-fired power stations tripped offline in a cascade. Load shedding started happening and even the New South Wales-Queensland interconnector tripped offline. But there was enough inertia in the system that it stabilised after about 15 seconds.

 In 2016 a number of South Australia wind power stations tripped offline in a cascade. The SA-Victoria Interconnector tripped offline, then there was not enough inertia, and the frequency collapsed so fast that system load shedding schemes couldn’t operate and everything protectively shut down in a state-wide blackout.

The article concludes by pointing out that all this unreliability has been purchased with increased power prices too! More money paid by consumers for a shittier system.

The following graph is not from Chris’s article but from my 2022 post, Lessons for National from the Australian Power Crisis

========

Australia’s Transition to Renewable Energy:
Part 1
Part 2