Six flaws in energy transition planning

The Australian Government wants to transition our electricity system to net zero carbon emissions by relying on wind and solar power — excluding alternatives such as nuclear energy. Its justification for this switch rests on the claim that a renewables-dominated system is environmentally beneficial and economically superior to using fossil fuels or nuclear energy.

But this premise is based on a flawed argument, which has left out key elements that portray a different picture.

Energy Minister Chris Bowen, policymakers and other public figures use the CSIRO’s GenCost report and the Australian Energy Market Operator’s (AEMO) Integrated System Plan (ISP) to support their argument. These reports defend a belief in the economic superiority of renewable energy which relies on two distinct claims:

1. That renewable energy is cheaper than alternatives, including fossil fuels, regardless of any cost of carbon or other policy constraints; and
2. That the planned transition is the cheapest pathway to reach Australia’s emission reduction targets.

The first claim is principally supported by GenCost, and the second by the ISP, though these claims are frequently conflated in policy discussions. However, the Centre for Independent Studies (CIS) has identified many critical flaws in the analysis of both reports that destroy any credible support they could give to these claims.

CIS research has found that using coalfired generation for the majority of baseload demand, which Australia is dismantling, would be significantly cheaper in the near term than relying on intermittent renewables for 90% or more of energy generated, as the ISP and GenCost do. CIS has also found that GenCost overstates the cost of nuclear plants because it does not consider cost-effective and established models.

Second, CIS found the ISP has been undermined in its aim to develop the lowest cost system by finding the optimal combination of infrastructure across states. State and Federal governments ‘declare’ projects to be required, without a comprehensive system-wide business case. The ISP then treats these projects as ‘sunk’ and forces the system to optimise around them, increasing costs. If the most inexpensive system was built — even relying almost entirely on renewables — it would likely look markedly different to the system currently being built. In particular, it would exclude very deep pumpedhydro storage (e.g. Snowy 2.0) and the rapid advancement of high-voltage, interstate transmission which is underway.

This paper focuses on the question of how Australians have been led to believe two false claims: that renewables are the cheapest and that the current plan is the cheapest version of a renewables-dominated system.

To answer this question, CIS has categorised the flaws in GenCost and the ISP under six headings. These flaws hide the true costs of the proposed energy transition.

The six fundamental flaws of GenCost and the ISP are:

Outscoping costs

a. GenCost excludes storage and transmission costs incurred before 2030, making wind and solar appear cheaper than coal, gas and nuclear, even with high wind and solar penetration;

b. The ISP relies on rooftop solar and home batteries to provide generation and storage but excludes their costs from the model;

c. The ISP excludes the cost of recycling wind turbines, solar panels and batteries, making renewables appear cheaper than they are over their lifetime.

Outscoping carbon

a. The ISP excludes emissions from the manufacture of wind, solar and batteries, making them seem cleaner. This restriction of ‘Scope 1’ emissions means the ISP will increasingly export emissions to China, whilst creating the appearance of meeting net zero ambitions locally.

‘Match-fixing’ the energy transition

a. The ISP claims a renewablesdominated grid is the cheapest option for Australia’s energy system without comparing any alternative, effectively fixing the outcome from the beginning, as the only baseline scenario without a binding renewable energy or carbon target has been removed;

b. The ISP manipulated the weighting of scenarios with more ambitious renewables targets to force a faster timeline for transmission projects.

Overfitting the model

a. The ISP has an overfit model that assumes perfect foresight of the weather decades in advance and builds just-in-time flexible gas capacity before years predicted to have poor weather for renewables. Overfitting occurs when models conform too closely to a limited set of inputs and fail to account for the variability of the real world. In reality, the grid will have to be prepared for almost any weather, every year, requiring greater investment to ensure reliability.

Disintegrating the integrated system

a. The ISP method for determining the value of individual projects does not treat the energy system as an integrated whole (i.e. a system of smaller sub-systems) but rather a collection of parts largely independent of one another, allowing uneconomic projects to be approved and costs passed onto consumers;

b. The ISP treats governmentcommitted projects with costs yet to be sunk as locked in and doesnot assess their benefits, making transmission projects that link these assets seem more valuable.

Cherry-picking data

a. GenCost cherry-picks a single, overestimated data point from a cancelled project to use as the cost estimate for Small Modular Reactors (SMRs) and does not include any data from large-scale, nuclear plants — making nuclear energy seem more expensive;

b. GenCost cherry-picks coal and gas price estimates so that fuel price spikes induced by the Ukraine war are locked in for the lifetime of new coal and gas plants;

c. GenCost uses unrealistic assumptions about construction costs for new coal plants, reinforcing the illusion they are more expensive than renewables;

d. The ISP cherry-picks certain future years in which transmission projects show greater benefits in their costbenefit analysis to justify projects going ahead as soon as possible in an entirely different year;

e. The ISP cherry-picks years in its reliability analysis in which no reliability breaches occur and ignores years when a breach is likely to occur.