The Grattan institute’s latest report Go For Net Zero, attempts to find a middle path between competing visions of the future electricity sector. On the one hand are certain Coalition politicians, who still believe that because Australia is largely powered by coal today, it must be best to keep powering it with coal into the future (regardless of emissions). On the other hand are the Greens who demand ever tighter timelines for full decarbonisation and then seek to constrain the technologies we might use to get there. The report politely disagrees with both, making the case for a net zero approach based around renewables with some back-up gas whose emission are offset as the next stage of evolution of the power system.
Perhaps wisely, the report doesn’t get too hung up on specific timelines nor on an overly detailed modelling approach that risk promising spurious accuracy. Instead, they carry out some high level modelling to illustrate a number of fairly common sense conclusions:
- Higher levels of renewables can reduce emissions at low cost
- There is no benefit (i.e., it’s not cheaper) to replacing old coal plants with new coal plants
- Targeting net zero emissions is cheaper than targeting zero emissions
- Targeting net zero emissions is cheaper than targeting 100 per cent renewables
- Net zero ultimately means finding emissions sinks that can extract CO2 from the atmosphere
- Gas can provide an economic backup to renewables for those rare periods where there is little wind or sun available for several days in a row
Grattan estimate that a mix of either 70 or 90 per cent renewables is cost-competitive with a coal-based system with the benefit of greatly reduced emissions.
Figure 1: System cost comparisons
Both the 70 and 90 per cent scenarios are also cheaper than 100 per cent renewables, even after allowing for the cost of offsets.
Grattan’s conclusions that gas still has a role to play conflicts with that of another new report, this time from the Clean Energy council (CEC). This argues that batteries are already cheaper, on either a $/Kw or $/KWh basis than gas peakers. Thus, they say there is no need for any more gas plant – batteries are the answer to everything.
Figure 2: LCOE batteries and gas peaker
One reason for the different conclusion is that the two are looking at alternative dispatchable plant options in different ways, The CEC does a straight levelized cost of electricity comparison between a generic gas peaker and two battery types: a four hour battery and a two hour battery (with the number of hours being the time it takes to fully discharge at maximum discharge rates). The assumptions are not unreasonable: a gas price of $6.50/GJ, 10% utilisation for the peaker, an average $30/MWh to charge the battery, etc. But the outcomes are somewhat dependent on the assumptions, as always.
Grattan is instead thinking about the need to cover “Dunkelflaute” conditions that can persist for several days where wind and solar output are low. Their demand and weather modelling is based on AEMO’s wh ich in turn are based on actual weather conditions over the previous decade.
Figure 3: Average wind and solar supply minus demand (GW), across different
There is a fortnight in winter 2013, where, if these conditions were repeated would require 9GW of dispatchable electricity for 14 days. They estimate this would be equivalent to 9 Snowy 2.0s. Neither overbuilding renewables nor further increasing transmission (beyond the substantial additions already assumed) is likely to be cost-effective. There probably isn’t enough scope to expand the snowy and Tasmania hydro schemes enough (and move the electricity around). Demand response is not going to persist for 14 days, and other zero emissions dispatchable technologies have their own challenges. Green hydrogen, for example, would require even more renewables to create and would need extensive storage to be available for when it’s needed, so gas is the best bet. Why not batteries. Well, even with four hour batteries, the NEM would need 6 for every day of the fortnight. 6 x 14 x 9GW = 756 GW of batteries. This doesn’t seem plausible. Plus, many of these batteries would have limited opportunity to make money at other times. This shows the flaws in just trying to think of options as abstract units of capacity. It’s not to say the Grattan is necessarily right and the CEC is wrong, just that we need to be thinking in terms of systems rather than individual technology choices.