Demand forecasting is likely to become more important and more difficult as the network transforms. It will be particularly important for efficient and timely investment in ‘lumpy’ infrastructure assets, such as new generation or network infrastructure.
Under-estimation of demand may contribute to under-investment, possibly resulting in security and reliability issues, while over-estimation of demand may contribute to inefficient and wasteful over-investment that drives up costs for consumers. The Panel has heard that recent investments in network assets were informed by inaccurate forecasts, which contributed to the networks being overbuilt, with consequent impacts on electricity bills.
Credible demand forecasts are essential to the operation of the NEM. Demand forecasts underpin all planning and investment decisions. Better informed demand forecasts reduce the likelihood of incorrect or inadequate responses to emerging risks and opportunities. Large demand forecasting errors have contributed to a number of problems in the NEM.
From the commencement of the NEM until the late 2000s, electricity demand increased with population and economic growth. From the late 2000s, demand moderated and then declined. Electricity forecasters did not anticipate the structural shift in electricity demand that would occur as a result of increased penetration of rooftop solar photovoltaic, changing industrial consumption, consumers’ responses to rising electricity prices, energy efficiency initiatives, and changes in the economic environment.298 Further, it took several years to recognise that the changes in actual demand were part of a structural shift. This is demonstrated in Figure 5.1.
AEMO has established a work program to improve the accuracy of its long-term demand forecasts, including moving from ‘top-down’ to ‘bottom-up’ models with finer resolution within regions, increasing stakeholder engagement, increased use of meter data, and increasing transparency of data, models and advice.299 AEMO is currently establishing a new performance assessment framework for its forecasts.300 The Panel supports this initiative to drive continuing improvements over time.
Collaboration with the CSIRO as it develops its Energy Use Data Model (EUDM) is likely to further strengthen AEMO’s modelling capabilities. The EUDM will provide detailed information about changing electricity consumption patterns, including peak loads and daily load shapes, and demographic, technological, and environmental factors interact to shape behaviour.
As the NEM evolves, there is a risk that existing network assets will no longer be required or will not be required to the same degree, due to the closure of existing generators or consumers.
This possibility raises the question of how stranded assets should be handled. Some stakeholders have called for the value of network assets to be written down, either voluntarily or compulsorily, where there has been over-investment in capacity to meet peak demand or major structural changes to the load.301, 302 Network service providers are opposed to compulsory write-downs.303
Voluntary write-downs of asset values would have the effect of lowering prices for consumers, and
reducing the size of the regulated asset base without increasing the risk profile of the investment. Compulsory write-downs are problematic. Writing down the asset values would increase creditors’ perceptions of risk, resulting in a higher Weighted Average Cost of Capital for future projects or refinancing, leading to potentially higher costs for consumers over all. The issue of historical network over-investment is beyond the scope of this Review, and it is not possible to examine this issue in detail. However, given that there is a reasonable likelihood that some assets will be stranded in the future, there is merit in a future examination of the most appropriate method for handling this issue in the long-term interests of consumers.
The uptake of new technologies is putting residential, commercial and industrial consumers at the centre of the electricity market. Distributed energy resources, such as rooftop solar photovoltaic and battery storage systems installed at commercial and residential premises, energy efficiency improvements, and demand response by consumers can all be harnessed to improve the reliability and security of the electricity system. Consumers of all sizes should be rewarded for taking those actions, which will benefit them individually and also help reduce overall system costs, leading to savings for all consumers.
Achieving this outcome requires action. The retail electricity market must operate effectively and serve consumers’ interests. Improved access to data is needed to assist consumers, service providers, system operators and policy makers. Increased use of demand response and changes to the role of networks and how they are incentivised are required to unlock these benefits. Governments also need to take steps to ensure that all consumers, including low income consumers, are able to share in the benefits of new technologies and improved energy efficiency.