An Introduction to Operational Forecasting at AEMO

Forecasting plays a critical role in the management and operation of a power system, and for AEMO it is the Operational Forecasting (OPF) team that provide operational forecast capability for both the National Electricity Market (NEM) and Wholesale Electricity Market (WEM). This article provides an overview of the functions provided by the team that go into delivering this important capability for our national energy markets.

As the team’s name suggests, a primary function of the team is to provide operational forecasts for the NEM and WEM.

In the NEM, the team provides forecasts for operational demand, along with distributed PV (DPV) and variable renewable energy (VRE) generation (for each individual large-scale wind and solar site). We also produce the Forecast Uncertainty Measure (FUM). The FUM is used in conjunction with the largest and second largest contingency in NEM regions to determine lack of reserve (LOR) triggers. There is a slight difference in the WEM, in that VRE sites are required under the rules to self-forecast, while the FUM isn’t produced.

These forecasts cover horizons ranging from 5-minutes ahead through to 7-days ahead and are critical inputs for power system dispatch, market efficiencies with 5-minute pre-dispatch and pre-dispatch) and operational planning, including pre-dispatch and short-term projected assessment of system adequacy (PD and ST PASA). The majority of this data can be found here for the NEM and here for the WEM.

The team is responsible for processing millions of forecast data points on a daily basis. For example, the Australian Solar Energy Forecasting System phase 2 (ASEFS2), which produces near real-time estimates and forecasts of solar generation for distributed PV systems, defined as systems with less than 100 kilowatts of capacity, generates approximately 531,000 data points per day.

The OPF team sits in the Near-Time Operations Department alongside other teams that work on planning, constraints and market dispatch, within the Operations Division of AEMO.  In comparison, the long-term forecasting teams in the System Design Division of AEMO provide forecasts covering many years for publications such as the Electricity Statement of Opportunities (ESOO), Gas Statement of Opportunities (GSOO) and Integrated System Plan (ISP). The OPF team consists of around 20 geographically dispersed members from across Australia, each with different backgrounds, such as data analysis, meteorology, data and software engineering, and renewable energy.

Operational forecasting is not just about producing forecasts. There are other important functions the team provides to AEMO and the wider energy industry. They include:

• Situational awareness for control room, operations planning and wider AEMO – this includes working with meteorologists to track severe weather and communicating potential risks to the NEM and WEM control rooms so operators can respond appropriately. Recently, this has been expanded to include space weather risks.
• Supporting AEMO’s onboarding and connections process for new assets – OPF plays a key role in the registration process for new wind and solar farms in the NEM, particularly through its review and approval of Energy Conversion Models (ECMs), which is mandatory under NER 3.7B. The team validates ECM content, checks SCADA signal consistency, and supports the setup of generation forecasting models in AWEFS and ASEFS for Dispatch, Pre-dispatch, and STPASA. Onboarding also includes assessing non-scheduled registration applications to determine whether forecasting is required, ensuring alignment with market and system needs. Additionally, the team integrates wholesale demand response units into the forecasting systems, and onboarding self-forecasting providers so participants can submit their own dispatch forecasts.
• Reporting and documentation – support AEMO’s reporting obligations, including providing forecast information for power system incident reports and contributing to the development and review of operating procedures related to forecast management for the control room. A new example of the team’s reporting obligation is the Unscheduled Price Responsive Resources (UPRR) Monitoring and Reporting portal. The portal evaluates the effect of active UPRR on the accuracy of AEMO’s short-term demand forecasts, in accordance with NER 3.10C.2(e)-(g).
• Ongoing improvement of tools and processes – maintain and improve our forecasting models by updating inputs, parameters, and variables to reflect system changes and improve performance. In parallel, we develop new tools to help monitor and assess the impact of emerging challenges – such as rooftop solar PV variability, sudden weather changes, demand-side response, and price-driven behaviour. Supplementing the forecasts that flow through to market systems, the team also develops several internal forecasts/tools to assist with forecast performance monitoring, decision support and situational awareness. These forecasts and tools utilise the team’s highly developed data science capability.
• Provide input to market reform initiatives – provide specialised advice from an operational forecasting perspective.

A key part of the ongoing success of the team is engagement with its internal and external stakeholders. The internal stakeholders, such as the control room, operational planning and connections teams are mostly reflected in the list above. Some of the team’s key external stakeholders include:

• The weather industry – provide weather forecasts and specialist advice on uncertainty and sudden changes in the weather that may impact the energy system. For example, the timing of a cool change. The team’s relationship with weather forecast providers is increasingly becoming more important as the uptake of renewable energy continues to increase.
• Energy forecasting vendors (demand and VRE) – provide demand and VRE forecasts to supply alternative views/scenarios or outcomes that help inform operational risk assessments.
• Other external data providers – the team utilises other sources of data to complement/supplement internal AEMO data sources. For example, data from the small-scale Renewable Energy Scheme (SRES) from the Clean Energy Regulator (CER), and generation samples of individual distributed PV systems from external vendors are used in ASEFS2.
• Market participants – to understand their needs and share risks that may impact generation. For example, high temperature cut-outs from wind turbines during extreme hot temperatures.
• Network service providers – liaise with the OPF team on actions that may assist in managing operational demand or network constraints
• The AER and AEMC – engaging with these to meet existing regulatory obligations and deliver regulatory change, where this affects operational forecasting.
• Software and platform providers – to ensure the team stays up to date with the latest digital technologies to provide the best forecasts/insights possible while ensuring cyber security.
• International system operators – regularly engage with other system operators to share knowledge, discuss emerging challenges and learn best practices in operational forecasting. AEMO is a founding member of the International System Operator Network (ISON), formed in 2024 which also includes CAISO (California), EirGrid (Ireland), Energinet (Denmark), ERCOT (Texas) and NESO (UK).

These relationships are important to help the team stay connected with the broader energy sector and to identify emerging risks that may impact the energy system, and the team’s functions. Through these engagements, OPF also provides transparency of its forecasts and functions.

Please keep an eye out for future articles to learn more about current projects and emerging challenges the team is navigating with the energy transition. Many of these projects and challenges arise from the continuing uptake of CER technologies and the increasing role these technologies, along with weather, are having on the generation mix and demand.

About our Guest Author

Dr Adrian Grantham has been working in the energy sector for over 10 years, including the past seven years in long-term Forecasting and more recently in Operational Forecasting for the AEMO. In long-term Forecasting, Adrian led the team responsible for developing, delivering and advising on medium to long-term demand forecasts of electricity and gas. In Operational Forecasting, Adrian is currently providing specialist advice on several challenges relating to the energy transition with a particular emphasis on consumer energy resources. Adrian is passionate about the modernisation of Australia’s energy networks and markets and enjoys working on the exciting and challenging opportunities the energy transition is presenting. You can find Adrian on LinkedIn here.