As part of my research reviewing all 100 % renewable electricity studies, I am revisiting the work of ClimateWorks for the Deep Decarbonisation Pathways Project. You can find the report here.
The report includes three scenarios:
- 100 % renewable grid (of interest to my meta-review, and the major scenario presented by ClimateWorks)
In the latter two scenarios, “permission” is effectively given in the model for those energy sources to compete. As a result, each of these energy sources comes to make a significant contribution to the least-cost finding (23 % and 14 % respectively).
The limitation on nuclear to only 14 % seemed surprising to me. It is based (primarily) on the following assumptions:
- Only a few large nuclear plants can be accommodated in the eastern states before connection limitations are met
- Western Australia remains separate from the NEM (despite the scenario suggesting it will become the largest electricity user)
On the first point they say “ESM (CSIRO’s Energy Sector Model) only allows nuclear power in the larger states of New South Wales, Queensland and Victoria, since nuclear power stations generally have to be a minimum size, which would preclude their deployment in the smaller states”.
The authors acknowledge that consideration of smaller reactors is important and worthy of further work. They also tested the WA connection in the nuclear scenario and found the nuclear share increased to 27 %. That alone is an interesting signal.
They state further that “with the use of small-scale nuclear plant or a more interconnected grid [we will return to that] there is no technical reason why nuclear power could not supply a major share of electricity consumption”.
In light of these sensible remarks, I am concerned that the assumptions about transmission network limitations and may not have been understood and applied equally. I am not certain of my findings here. All of the following is to be read as open investigation, not conclusion.
The authors state:
“Within the scope of this project, it is not possible to fully resolve the changes to the network that would be required to support the reliable market balancing of the generation mix in each scenario. The relatively high demand growth will necessitate significant transmission network growth and the need to extend the grid to remote renewable resources, given their high contribution to the generation mix”.
Ok… they know it’s essential, it was out of scope. Fine. I have no problem with that. What did they do?
“As a guide, in the absence of detailed transmission network modelling, the transmission cost results from the existing ‘Renewables thrive’ scenario of the Future Grid Forum project have been applied, which is a 100 percent renewable grid scenario, as outlined in Graham et al. (2013)… Under ‘Renewables thrive’, taking into account the required changes to the transmission network, the cost was projected to increase to 4.3c/kWh by 2050”.
So, no modelling, but an effort at costing based on someone else’s modelling. That, again, is fine with me as long as it’s disclosed.
Now, let’s look at the mix for 100 % renewable grid.
It is evident that this mix will require the “significant transmission network growth and the need to extend the grid to remote renewable resources” that the authors discussed. So, I strongly suspect the transmission was assumed to be unconstrained, simply costed in on a per kWh rate on the basis of “we will build what we need”.
Here are the network augmentations proposed by Future Grid Forum.
That’s a lot of investment that could be put towards improvements tailored for nuclear generation.
So…why was transmission assumed to be constrained in the nuclear scenarios such that only a few reactors in the eastern states were assumed to be possible? Renewable units are smaller than nuclear units, but that renewable scenario is (much) more dispersed than a nuclear-based system would be. Stronger transmission networks (and other investments) is an answer to both problems. The authors acknowledge that very fact (“with the use of small-scale nuclear plant or a more interconnected grid there is no technical reason why nuclear power could not supply a major share of electricity consumption”).
Every progressive transmission enhancement that would be necessary for the 100 % renewable grid scenario would, just as effectively, contribute to new threshold that enables the connection of more large nuclear units. Has this modelling settled for assuming nuclear connections based on the network of today, but enabled levels of renewable penetration based on the network of “whatever we need tomorrow?”. If so then an assumption that should be uniform across all scenarios is, instead, differentiated between scenarios.
The transmission enhancement requirements for nuclear (which can be close to load and existing transmission) are likely far more discreet than for the needs of the renewables system shown above. I posit dollars invested in transmission upgrade for nuclear would deliver connection capability much more efficiently than dollars spent in transmission upgrade for geographically dispersed renewable overbuild. If spending on networks was applied rationally and even-handedly, nuclear should do well.
The result is that the contribution from nuclear in the scenario where it was permitted would be, at 14 %, a false low. This suggests that the wholesale price of the nuclear scenario (Figure 2.18) is a false high, as more nuclear, earlier, would displace more of the higher-priced renewable electricity that enters the mix later: enhanced geothermal and wave (capital costs, Figure 2.7) and solar thermal (which, while dispatchable, is still variable). It’s likely not delivering an optimised system from the available options.
I find this report is overall clear, well written and offers some important potential directions, as well as giving one of the more even-handed treatments of nuclear generation in Australian literature (my criticisms not withstanding). However I cannot get clarity on this issue. I welcome comments from all including the report authors if they can shed some light on how this assumption was managed.