Back in the early 2000s, cost reduction was the order of the day for energy trading and risk management (ETRM) technology systems. Risk management around energy price was becoming a fringe-boardroom issue, thanks to mounting commodity volatility and ETRM specialists would hedge bets against weather patterns and other issues in order to shrink the company-wide fuel bill. Just a decade later, energy is on the boardroom agenda in a much bolder way. Real-time trading across multiple energy markets, as provided by Limejump, is now a viable option for energy generators, bringing new revenue opportunities to traders of all types and supporting the UK’s decarbonization efforts in ways the 2000s trader could hardly have imagined.
In this piece, Limejump’s Genna Boyle and Luke Peck explore how this evolving landscape is being driven by technological advancements in artificial intelligence and machine learning.
Saying farewell to coal
As the UK phases out coal-fired power generation—before 2025—in a bid to fight climate change, the role for intermittent renewable generation is rapidly growing and National Grid requires more flexible, on-demand power generation to ramp and deliver. At the same time, demand profiles in the power sector are shifting, caused by a myriad of lifestyle demand changes, from time of use tariff, electric vehicle charging, and new heating initiatives. The demand patterns of old, with clear morning and evening peaks, are changing as behavior evolves.
"Algorithmic decision-making frees up time for high-value decision-making in complex trades where the best returns can be found"
This has paved the way for the modern balancing mechanism (BM) for short term grid balancing actions that protect the UK infrastructure from demand surges. This enables battery storage and other standby fast-responding assets to play a major role in enabling more renewables to go online and mitigating short term fluctuations in the frequency of the power system. This is where modern traders can add real value, to both asset owners and National Grid, by offering flexible and fast ramping megawatts in real time.
Rise of the machines
The growing percentage of renewables in our energy mix—evidenced in recent coal-free periods—and changes in national demand profiles bring more volatility which can erode profit margins due to forecasting difficulty.At the same time, while the rate of technological development in renewable assets has declined, there have been major changes to the way the grid connects and the components “talk” to each other, increasing the requirement for old generation tech to communicate with modern data processing.
In this new landscape, forecasting becomes crucial. Power pricing is influenced by all sides: Brexit, the global commodity trade, and geopolitical actions alike cause fluctuations. Accurate forecasting mapped against weather patterns help mitigate negative effects of this influence by more accurately matching supply and demand—and in turn, supporting traders to achieve revenue. Smart traders are also making the most of assets that can play in multiple markets and “stack” to create healthy returns.
With smaller, more complex margins spread across numerous markets, human forecasting becomes less accurate and compromises returns, and this is where the digital data era is coming into its own. Artificial Intelligence (AI) and machine learning are now being used to inform algorithmic trading decisions, which in turn are automatically utilized in ETRM systems, thus reducing the reliance on the manual workload of traders. Algorithmic decision-making frees up time for high-value decision-making in complex trades where the best returns can be found. Fast data processing feeds machine learning to sharper weather and pricing forecasts that provide traders an edge with real-time data, tracking everything from the Net Imbalance Volume (NIV) to current power import and export status.
With the markets evolving at an unprecedented pace and new technological advancements taking place every day, the use of predictive analytics and machine learning increase. We anticipate that AI will become progressively pertinent with IOT connected devices—from washing machines to EV chargers—using these to interact with real-time power markets. This sector moves quickly. Overnight, entire products have paused—like the Capacity Markets—and the assets we trade, including batteries, solar panels, wind turbines, and more, are being utilized in markets that were never designed to use them.
The next steps for this industry will develop as it embraces data skillsets as critical tools, not modern science fiction. For companies wanting to make the most of this technology, they need to select partners that are prepared to walk before they run: understanding the subtle ins and outs of power balancing and trading before technical programs are developed. With both industry knowledge and computing behind them, traders can truly harness the power of machines in modern energy trading, and maximize the financial returns on offer.
Forecasting for the future
Modernization isn’t happening in the trading corner of the energy industry alone. The rise of EVs, the move to electric-based heating, and increased action to fight against climate change, not to mention infrastructural changes such as the boom in data center needs, are contributing to an unprecedented electricity demand escalation and drastic changes to traditional demand patterns and peaks. AI and machine learning open up valuable tools to meet this demand, support national decarbonization, and capture the opportunity for revenue generation as the market evolves.