The mission of this group is to bring together utility professionals in the power industry who are in the thick of the digital utility transformation. 

Post

Energy Central Power Perspectives™: Autonomous Operational Forecasting, an Interview with Frank Kreuwel of Alliander

image credit: Esmee Hulst

Being able to predict the future used to be a magic trick, but today it’s a necessary reality for stakeholders across the energy sector. From predictive maintenance to anticipating demand, being forward-looking is not a luxury but a requirement for an efficient and optimal power grid. Luckily, advances in artificial intelligence, remote sensors, and other technologies have made accurate forecasting more and more possible.

As one of those experts who is advancing this critical field, Frank Kreuwel sees plenty of reason for excitement and innovation in the coming years. Frank is going to be sharing his insights at the upcoming virtual conference ‘AI&ML for the Smart Grid 2020’ with his presentation ‘Automonomous Operational Forecasting – leveraging AI-powered, short-term forecasting to enable autonomous local balancing of the grid while maintaining quality of data, models, and results under operational considerations.’

Ahead of his virtual presentation, Frank was kind enough to give the Energy Central a sneak peek for the Energy Central Digital Utility Community as a part of our Power Perspective™ Interview Series. If after reading you’re compelled to learn more about Frank’s work, be sure to register for the AI&ML for the Smart Grid 2020 virtual conference taking place on September 9.

On his background:

“I studied physics and astronomy in Nijmegen (NL) and I was really interested in renewable energy, as for me this is a field where fundamental knowledge can really be put into practice.  For Alliander, a Dutch DSO, the pace of the energy transition is a real challenge, the speed of change in use of the electricity grid outpaces the speed of grid reinforcements. Therefore, we require smart solutions which are able to make better use of the grid we already have in place. I really enjoy being part of a team to tackle issues with a strong societal relevance.”

 

On the Value of Forecasting

“Planning (and forecasting) is important on timescales ranging all the way from years ahead (should we build a new substation) down the shortest timescales where you can still take any practical action. Our team really focusses on the short term, from minutes to days ahead. Accurate insights in what will be the load on the grid for the next hours allow to really make the most of its capacity. For example, suppose a new solar park would cause grid congestion for one week of the year. Traditionally we would not be able to connect this park to the grid. However, using smart technologies, we can exactly pinpoint whether tomorrow will lead to congestion or not. And if it would, we can change the configuration of the grid, or use grid curtailment to prevent congestion, while allowing the PV park to feed-in its generated solar energy.”

 

On Automation vs. Human Intervention

“Human judgement is extremely important and will be for the next couple of years (or decades perhaps). For example, when configuring new mitigation tactics for congestion-areas, human domain experts define what is an acceptable level of risk, and of temporary grid overload. However, that is the ‘design’ phase, for the operational phase we already see that many decisions can be made by algorithms autonomously (does the forecasted load exceed a certain threshold, should the model be retrained, should an alarm be raised). In our vision on future grid management, there will still be a role for human operators, but as the world changes, they will rely more and more on information (pre)processed by automated algorithms.”

 

On Challenges Ahead in This Field

“A big hurdle is the quality and availability of (real time) data. Another is the forecasting of extreme events. By definition, these occur almost never, but many machine learning algorithms require a lot of representative training data. And a very big one is acceptance of failure. Forecasting is inherently inaccurate. We have to be real in what levels of accuracy can theoretically be achieved, which can be expected in practice, and which uncertainty will remain. It will be up to human judgement to determine of this is acceptable for a specific use case or not.”

 

On the Important Topics of Discussion to Expect at the AI&ML for the Smart Grid 2020 Conference

“There are a lot of interesting topics! But if I would have to choose, I’m really looking forward to hear how other parties are making use of their realtime data or smart-meter data as these are hot topics at Alliander as well.”

_____________________________________________

If you’re interested in hearing more about Frank’s insights into asset automation and forecasting, be sure to check out his presentation at the AI&ML for the Smart Grid 2020 virtual conference, taking place on September 9. You can check out the agenda and register for the conference here.

Frank Kreuwel's picture
Thank Frank for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member

Discussions

Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.

No discussions yet. Start a discussion below.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »