Occam’s Razor

It is time that we apply Occam’s Razor (the principle of solving problems and developing models with the simplest set of assumptions) and George Polya’s Principles outlined in his book “How to Solve It” to the Energy Transition.

The national debate is focused on the relative cost of generation from very large infrastructure investments of great complexity. It is a common experience that large, complex projects will take longer and cost many times the estimated expenditure. Many infrastructure costs are regulated assets with government-approved returns on assets that encourage higher levels of investment.

What is notable about the modelling (especially with Levelised Cost) is the boundary constraints, inappropriate selection of comparisons, changes in fundamental variables (like demand) over long time scales, externalities, and opportunity costs. 

Taking a cue from Polya, it is essential to understand the problem. This entails understanding the shape of demand as it evolves in the context of the supply and pricing of energy when required. Users of energy have agency. They can change the shape of their demand (including intertemporal instead of battery storage), improve the efficiency of energy use (as is being done with data centres), and develop distributed off-grid power sources. Business consumers of power might relocate by State or internationally to find cheaper and more reliable power. 

An underlying and most serious problem is the urge for Australia to lead the world in reaching net zero. That is where we should take careful note of precedents and implications. Germany is the obvious example where its green policy made it dependent upon imports of Russian gas and French nuclear power. It hollowed out German manufacturing and made Germany commercially and strategically vulnerable. Australia can't pace the transition ahead of its trading partners without adverse consequences and the counterproductive effects of importing embedded carbon.

In contrast, Texas has been among the lowest cost power systems in the USA, with a mix of sources for electricity generation being natural gas 45%, wind (a US leader) 25%, coal 10%, solar 10%, and nuclear 8%. The elements of a plan might be to examine closely how we might best manage demand to optimise power usage and respond to changing load profiles over time. The technology is available to enable better management. Then we should look to countries that have achieved low power prices and resilient networks on the journey to energy transition and draw on their experience. We might then contrast how greater emphasis might be placed on distributed networks and colocation of generation, as researched by the US National Renewable Energy Laboratory. Finally, in the long journey, we need to be adaptive and learn from emerging circumstances by following a Bayesian process.