AI-driven risk management for optimised energy tradingOpinion
Posted by: electime 30th July 2021
By Krishnan Kasiviswanathan, chief operating officer, Innowatts
Wholesale electricity price volatility continues to increase as a result and correlated to the unprecedented changes in weather patterns and renewables penetration. Energy traders and regulated utilities are experiencing an increasingly complex challenge in managing their market positions to maintain profitability and attain key performance metrics. One example from the retail energy side of the equation is the measurement of the mean absolute percentage error (MAPE). This is a measure the industry has become accustomed to leveraging when forecasting energy consumption and the related procurement of supply.
The fact is wholesale electricity price volatility is a risk that carries the potential to cost millions in pounds/dollars in the event of uncovered market positions – a reality that became all too real for several energy suppliers this year – and the consequences of which can be devasting.
Energy risk – The potential cost of risk mismanagement
In recent months, inadequate load forecasting and risk management has seen several energy suppliers over-exposed to volatile spot market prices. In February for example, severe weather caused ERCOT’s wholesale prices to soar to $9,500 (£6,735) per MWh. Following the event, Brazos Electric, alongside several other suppliers, filed for bankruptcy after the company received $2.1 billion (£1.49 billion) in unexpected bills from the grid operator.
The ability to leverage near-real time meter-level customer data and other external data resources through new, proven-at-scale AI-driven platforms, such as Innowatts, provides energy suppliers with the ability to enhance their risk management capabilities, reducing the potential for over-exposure. Alongside other related workflows such as load forecasting, meter-level insights and analytics it also better informs customer engagement and short- and long-term forecasting planning efforts.
The Innowatts’ platform integrates directly with any energy supplier’s energy trade and risk management systems to provide traders with a holistic view of all of their positions. This capability coupled with the integration of AI technologies provides traders with the option to run a variety of ‘what-if’ scenarios on the data. This provides a much more informed trading position that enables data-driven decisions that better assess the various generation sources available, and positions taken or contemplated.
While beneficial – optimise load forecasting as a first step
A critical first step to take before introducing more advanced risk management practices is to optimise load forecasting capabilities. Afterall, risk management insights are only as good as the load forecasting data that it resides upon. See Chart 1 below:
Optimising load forecasting does not necessarily equate to acquiring new data, nor a need for smart meter data, rather, the primary focus is on exploiting what is already available to seek out additional granularity.
There are multiple sources of data that can create greater transparency and offer energy suppliers valuable insights to the data that currently exists. For example, substation and feeder metrics, meter data, billing history, wholesale pricing, socioeconomic and other related demographic data, alongside generation and weather forecasts can all be used by the Innowatts platform to create more accurate load forecasts. This provides energy suppliers and traders the flexibility to build out the platform’s common data model, depending on the data that they are most interested in analysing.
For those energy suppliers that have a large portfolio of customers the AI driven algorithms can be executed on any customer segment(s) to reap the benefits of near real-time meter data. These options combined create a centralised forecasting analytical solution that is unprecedented.
A strategic and competitive edge
In an increasingly competitive and volatile landscape, accurate load forecasting and optimised risk management is an essential capability for all energy suppliers and vertically integrated utilities across the globe. With advanced data analytics, the most mature energy supplier trading desks will not only be competing to secure the best possible price and market position but have a mission to drive down the margin of error and increase profitability or rate payer returns.
Altogether, incremental improvements to risk management and load forecasting stand to save energy suppliers and vertically integrated utilities hundreds of thousands of pounds/dollars in their everyday business operations.
However, as the recent Texas blackouts, the heatwaves and shortages in supply across the UK and in the Pacific Northwest illuminate, it can take just one extended event to recoup the return on investment many times over by leveraging modern technologies and meter-level data-driven platforms.