If you have an interest in how Australia’s future grid will take shape, and the pace of the energy transition and emissions reductions, dare to cast your gaze away for a moment from the typically sordid politics taking place in Canberra.
There are now enough clues to suggest that the key decisions will not be taken only in the corridors of Canberra’s political elites, but also in the computations and assumptions of the energy wonks trying to model what that modern grid looks like.
And much of this is happening within the confines of Australia’s Energy Market Operator. The main game is not, as some would suggest, the reliability obligation of the proposed National Emissions Guarantee, but in AEMO’s own Integrated System Plan.
The ISP has been a work in progress for some six months now, and broke cover just before Christmas when AEMO boss Audrey Zibelman spoke of the need for a coordinated plan, and the need to model emissions reductions far beyond what was being contemplated by the current government.
The ISP will be a template that will be hugely influential in the key decisions to be made in coming years – the need (or not) for multiple new transmission lines, the need to focus on large-scale and distributed storage, and the choice of energy technology.
The first draft of the plan is not going to be unveiled for a month yet, but already there are some fascinating things taking place in the assumptions and scenarios that are to be fed into the machines that will do the modelling.
Last month we reported on AEMO’s hugely optimistic scenarios for the uptake of electric vehicles, suggesting that they could double over its previous assumptions made just six months earlier, and within two decades 10 million EVs could be on the road.
RenewEconomy has now seen the modelling that will underpin other key elements of the plan – including that for rooftop solar, small-scale batteries (and the extent to which they are linked), and the cost of competing technologies and fuels – wind, solar, gas and coal.
Some of the assumptions are quite stunning, others may be a cause for concern for certain parties.
Among the most interesting are the forecast falls in the cost of solar power, which the AEMO modelling suggests will fall another 50 per cent from a current LCOE (levellised cost of energy) of $64/MWh.
By 2036, it may cost just $34/MWh, a figure far lower than most other official modelling from the likes of the Finkel Review and others – although not cheaper than some forecasts from private analysts, and solar pioneers such as Dr Martin Green of UNSW.
The cost of battery storage is also modelled to fall dramatically – by almost 50 per cent to $133/MWh (calculated over two hours, single cycle), while solar thermal and storage technology costs are tipped to fall by one-third from $121/MWh to $81/MWh.
These three technologies are the big movers in the electricity market, and therefore the big influencers over future choices. Even wind energy is modelled to fall only slightly over the next 20 years, to $59/MW from $63/MWh, which may surprise a few.
All other technologies, from pumped hydro to gas generation are modelled to be static. Fuel costs for coal are expected to rise, the modelling does not appear to contemplate any new coal generation, which will be a surprise to no one, apart from the feigned horror of the far right.
The forecasts for rooftop solar, and batteries, and how many of those batteries are integrated into virtual power plants, are also interesting, as are the implications for the democratisation of energy and the impact on the big incumbent utilities.
The modelling assumes that rooftop solar – for both households and businesses – will continue to boom, and by 2050 will be providing between 35,000GWh and 60,000GWh of electricity per annum.
That equates to more than 20 per cent of supply in most states, and in Victoria – remarkably – it equates to 33 per cent of total demand.
In small-scale batteries, the modelling varies between 7,000MW and 20,000MW of battery storage by 2050.
Then it considers how many of these are actually connected, via virtual power plants, to provide arbitrage and emergency response – 10 per cent, 45 per cent, or 90 per cent.
In one of the scenarios, high distributed energy, battery storage accounts for nearly 40 per cent of demand.
The modelling then needs to dial in a range of other assumptions. Firming capacity, for instance. The grid design is focused on meeting peak demand, reasoning that what was once considered “baseload” will look after itself.
An interesting thing to note here is that solar goes from a 25 per cent firming capacity rating to zero, the more that is built.
The first wave of solar plants are credited with having reduced the day-time peaks, so had a rating of 25 per cent, but once more solar is built, the peak then moves to a time when solar has no influence – hence the zero rating for the later additions.
Other assumptions include lead times for new construction (renewables and batteries are by far the quickest), and then the big question about options to upgrade the grid – particularly new lines that link states or service renewable energy zones.
So, what does all this mean? Well, we don’t know yet. For that we will be closely analysing the speeches and presentations by AEMO boss Audrey Zibelman, and the actual Integrated System Plan when its first draft is released in June.
This, however, is likely to be only the start of the conversation, and one thing that can be sure, it will set off a set of vigorous debates.
For a start, apart from the price and other assumptions mentioned here, the modelling will take different overall scenarios into account – slow, neutral and fast change – as well as high DER, and other assumptions including possible life extensions for existing coal plants, “forcing” both Snowy 2.0 and Tasmania’s “battery of the nation”, and other sensitivities.
One of the biggest debates will be the apparent competition between building bigger grids and more links, and focusing on distributed generation.
This is turn pits demand-side solutions against supply side solutions, centralised facilities verses regional, the choice of storage (pumped hydro and battery storage), synchronous versus inverter technologies, and of course, fossil fuels versus renewables.
And this may very well be the final call. AEMO, and most other analysts, have now accepted that the grid is changing and will change further.
The uptake of solar, batteries, EVs, and the cost advantage of wind and solar will see to that. The question is how fast can this be done, and should be done?
The faster decarbonisation calculations will be fascinating. And ultimately, will be the assumptions that count the most in the modelling, and will influence just how many of those existing coal-fired generators can stay on line past their current use-by dates.
AEMO has taken it upon itself to model a 70 per cent emissions reduction by 2050 as the bare minimum, even though Australia has no long-term target (TonyAbbott got rid of that when he scrapped the carbon price, and a 90 per cent reduction (the Paris target).
