A new study by the Gas & Energy Transition Research Centre has shown that more robust assessment of weather uncertainty is required, otherwise long-term planning studies could substantially underestimate the future requirements for peak-day firming in the National Electricity Market (NEM).
As the NEM becomes increasingly reliant on supply from wind and solar generation, its vulnerability to weather variability will increase, as will its dependence on the infrastructure that provides ‘option-of-last-resort’ power generation. The new research demonstrates that systematic assessment of weather and other integrated uncertainties would be needed to effectively plan investment into gas-powered generation, gas supply and other infrastructure critical to maintaining NEM resilience.
The study, led by Dr Joe Lane, developed an 80-year dataset on wind and solar resource variability, aligned with the NEM’s renewable energy zones and calibrated to AEMO’s own (more limited) renewables ‘capacity factor’ data. The analysis shows this new dataset can be complementary to the published AEMO info, providing a new resource for electricity and gas sector stakeholders concerned with ensuring an effective and rapid transition of Australia’s energy system.
For more information about the study and its findings, contact Dr Joe Lane.
Study abstract
As Australia’s National Electricity Market (NEM) becomes increasingly reliant on supply from wind and solar generation, its vulnerability to future weather variability will increase. To assess that risk, much of the energy transition modelling community relies on data published by the market operator (AEMO), which at the time this work was undertaken, covered 13 years (2010-11 to 2022-23) of wind and solar capacity factor (CF) variability. Although other analyses have illustrated that time-period may not be sufficient to cover the low-side risk from variable renewable electricity (VRE) generation droughts, there have been no alternative datasets developed that maintain alignment with AEMO’s renewable energy planning zones.
This study addresses that gap, aiming to improve the application of weather-related uncertainty analysis within long-term transition pathway modelling. CF timeseries are produced for the NEM’s renewable energy zones (REZ), calibrated to the AEMO CF data over their 13-year period, then back- cast over a longer timeseries (80 years, at 30min resolution). This is repeated using weather data from each of the BARRA-R2 and ERA5 reanalysis products.
The modelled CF show clearly that AEMO’s CF data does not adequately capture the variability in VRE generation potential over that longer historical record. An example case study shows that, for each year of the transition through to 2044, relying solely on AEMO’s data could potentially underestimate the NEM requirements for peak-day firming by ~30-60%. If gas-powered generation (GPG) continues to evolve as the NEM’s primary ‘option-of-last-resort’ service provider, as is currently anticipated by multiple planning studies, underestimating the need for gas supply to GPG (by that margin) could threaten future NEM resilience.
Key strengths and weaknesses of these novel CF datasets are illustrated, aiming to improve their application in applied uncertainty analysis of energy system transitions. The methodology chosen for this study is capable of (a) being rapidly updated when future AEMO data is released; and (b) incorporating other methodological improvements that would expand the scope of use cases for which the data is relevant.
