ForTheDoers Blog

What does the future of physical power trading look like?

Martin Lindström 04 May 2023, 9:43 EEST

Computer screens with numbers, models, and detailed weather data. All with one goal: to produce and sell power to end users when they need it, as efficiently as possible. The rise in renewables is now changing the way we trade electricity – and sometime in the future, it might not be humans doing the trading.

Physical power traders at the trading desk

Every day, physical power traders use complex modelling and calculations to forecast power demand and prices and sell electricity for the next hour, day, week, and year. This is a complicated process that involves taking into account weather forecasts (Is it windy, sunny, or is it going to rain? Is it going to be a low demand day or a cold day?), possible outages at power plants, and projections of power demand by industry, among many, many other variables.

Some are also involved in dispatching. Dispatching refers to the decision to deliver power from a power plant to the electric grid. This requires monitoring and controlling multiple assets and deciding how much power is generated at a given time based on supply and demand. Like with trading, the time scales involved vary: from reacting to changes in the balance of supply and demand within one to several hours to forecasting how much power is needed the next day or over the next few weeks. Much of the response comes from hydropower. Thanks to hydropower’s ability to respond rapidly to shifts in the power market, dispatching can be intense; on top of this, you must always keep in mind permit conditions and safety.

Trading is exciting and often hectic. The amount of modelling and math involved means that many traders have backgrounds in engineering, economics, and math. To be successful, it’s important to have a trading mindset: you must be eager and curious to constantly learn what makes the market tick and how to optimise the portfolio against it. Much of this is about dealing with uncertainty, as every day is different: the wind can pick up, the rain doesn’t fall, or a power plant is suddenly shut down for maintenance. One positive is that you won’t have to wait long to get feedback on your positions and tactics, since you’ll find out if your assumptions in the forecast and position were right the next day or, in the case of longer forecasts, by next spring.

As a trader, you always need to be on your toes: since there’s always something on the horizon, previous models and ways of thinking always have to be reconsidered. Here’s one example: with many cutting down on power consumption, we’ve seen that the traditional “sauna peak” between 7 and 8 p.m. in Finland has largely disappeared. This used to be something you’d consider when projecting prices for that hour the next day; for the time being, however, it’s no longer relevant.

The electricity system is like a basketball team

The biggest change in trading has come from the increased deployment of renewables. A decade ago, wind used to be a novelty, but today it has a major role in determining prices, while hydro and nuclear help secure supply when the wind doesn’t blow.

The increasing share of intermittent production has made the market far more dynamic and fast-moving, as power supply swings quickly depending on the weather. As a result, the role of day-ahead trading is dropping; with intraday trading playing an ever-larger part, the speed at which you have to react to changes grows with every passing day. Projecting prices an hour ahead is challenging enough, but in the future, this will drop to 15 minutes ahead. At such a time scale, it is hard for humans to consider all the necessary variables before making decisions.

So, how might trading evolve? Drawing on my experience from playing basketball, I see striking similarities between power markets and team sports. As with a basketball team, every actor in the power market has their specific roles and responsibilities. In my time on the court, I played the position of power forward, which required strength, versatility, and a high success rate when scoring. But I also needed to be part of the team and take on my role as a defender while clearing space for teammates to score. In basketball, success is based on having a team with the best fit and combination of capabilities on the court at the right time. Similarly, in future power markets, the five strongest players include hydropower, which will play a key role in meeting the short- and long-term needs of the power system, along with nuclear, wind, solar power, demand and response solutions, and – as a potential sixth player – hydrogen.

New rookies coming on?

Excitingly, there are some new rookies entering the game: machine learning, data mining, algorithmic trading, and AI. As I reflect on my own experience with a new team player joining our squad, I recall feeling both excited and sceptical. Would they strengthen our team, or would they just create chaos without any real results? Could they even take my spot and reduce my time on the court?

Similarly, I wonder how these new capabilities will impact physical power trading. As ChatGPT has demonstrated, in certain cases AI can write better text or computer code than humans. Likewise, AI can respond more effectively to rapid shifts in electricity supply and demand, while ensuring that power plants are utilised optimally and trades are optimised.

But just as us humans will still check what AI has written, human traders will still need to set the AI’s parameters and respond to any hiccups. This will require more and more traders with skills in coding and programming to run the AI and understand what it’s doing. Ultimately – whether power comes from your neighbour’s electric car or is sent to you by a neural net-leveraging trader – electricity is still generated and transmitted through physical assets. This means they can (and occasionally will) fail. In these situations, physical traders will be around to make sure that everything runs smoothly.

Nevertheless, it's clear that there are new kids in town!

Martin Lindström

Head of Nordic Optimisation & Physical Trading

Blog
To the blog frontpage