Analyzing and forecasting electricity distribution networks – part 1

By Jonas Jacobsson

Production and consumption of energy have to go fossil free, and at the same time, almost everyone and everything are consuming a lot of energy. Much is being done to help consumers decrease this consumption, including using LED for illumination, recommending better insulation in buildings, and providing a host of devices suggesting for optimal energy use. Nevertheless, in industries, transportation and homes – the quantities of consumed energy are rather high. To reach the global goals of affordable and clean energy, a lot has to change. Specifically, there needs to be a shift towards fossil free electric energy. Is it possible for us to make these changes less drastic by using existing infrastructure, providing that we use it in a smarter way? Let us consider what solutions already are here for electricity distribution companies, and ready to be utilized.

How can we use existing electrical distribution networks better?

One of the greatest challenges we face today is that the electricity distribution networks were not built for the currently increasing consumption. Building new electricity distribution networks is a cumbersome process, which includes planning and investments for up to 50 years. So, what to do? At Digpro, our Network Information System (NIS) dpPower, has tools that ensure that existing electricity networks are used in a smart way. It all comes down to balancing production and consumption, thus preventing overload and ensuring an electric network that has a levelled grid. We consider short- mid- and long-term forecasting, and how these are applicable as regards to electricity production and consumption. By establishing this, we can continue to use the electricity distribution networks we already have.

Digital Twin
To analyze the electricity distribution network, a Digital Twin supports real-time operations, planned operations as well as investment decisions.

Short- and mid-term forecasting

Short-term forecasting refers to what is happening now, within one hour or mostly two days. Mid-term forecasting brings us up to between one day and a month. An electricity distribution network is volatile, since it often relies on weather dependent sources when producing energy, as well as irregular consumer consumption. To simplify, one can say that:

  • Electricity production – relates to weather dependent energy, including lack or excess of sun and wind, as well as temperatures which often vary due to seasons. During the cold seasons, more energy is typically used for warming up houses, whilst less sun related energy is produced.
  • Electricity consumption – where industries, that are huge consumers of energy, typically consume more during weekdays. Domestic consumption tends to peak during evenings and weekends. Due to an increase in price volatility, the electricity consumption is also more volatile and unpredictable.
Electricity is used in many aspects of life, and electricity distributors benefit from knowing the peaks and lows of consumption.

Forecasting solutions have been around for years, and new tools such as machine learning and AI are now readily available. Then again, forecasting is to be executed on medium or even low voltage levels. This will greatly increase data volumes, therefore making the more advanced forecasting tools impractical. The solution is to find the correct tool for high volume data. At Digpro, the way forward is to use existing tools for network calculation, and the combining of existing load profile data with real-time measurements. Then, one will not need to do advanced integration on network models, and still provide a forecast down to low voltage network.

Does this work? Indeed, it does. Still, more practical experience from the live network is needed. And at Digpro, we work closely with our customers to ensure that our solutions meet the requirements of the electricity distribution market.

To learn more about long-term forecasting, please continue to read the second part of this blog.

Meet the expert
Jonas Jacobsson is CEO at Digpro Solutions AB, and has numerous years’ experience in the power and electric industry. He holds a Master of Science in Industrial Engineering and Jonas Jacobsson Management from Linköping’s University, and a Master of Science in System and Control Engineering from Case Western Reserve University, Cleveland USA. After a trainee position at ABB in 1996, Jonas continued his career focusing on IT solutions for electric distributions. He has worked at Digpro for over thirteen years, focusing on sales and management.

 

Sign up for our newsletter!

Enjoy the latest information about what is going on at Digpro – read about our products, news, events and much more.