Although pretty much anyone can be involved in IT today, Jan Konrád from Unicorn believes there are few actually capable of working in this field and solving problems related to energy transformation. That’s why it’s important to set up a solution that uses as much of what already exists as possible. At the same time, IT must handle consumption and production forecasting, energy balancing, smart metering, cybersecurity, and so much more. “The price of forecast deviations is growing, and whoever is better at forecasting will earn the most,” says Jan Konrád during his interview.
What do you see as the most current topic in the energy industry right now?
There are loads of changes in the sector and it isn’t easy to point to only one of them. But either way, before too long, we’re going to have a completely new energy industry here than we did 20 years ago. Building the sector as we know it today – let’s call it the “traditional energy industry” – took about 50 years. The transformation into the new energy sector won’t take as long, but it’s also not going to happen overnight.
The changes currently underway are rather dynamic, and we’re all already starting to feel the consequences. It definitely doesn’t look like we’re going to have access to an unlimited amount of relatively cheap energy, like we’ve been used to in the past. However, the transformation will also bring new opportunities, with many of them involving IT, which is a positive for our sector. For energy consumers, on the other hand, this isn’t such a positive thing.
What will these new solutions and products require?
It’s not just about the product but also their implementation. For instance, when we look at the expected investments into transmission and distribution networks, the curve takes a steep spike upwards. Compared to previous years, we’re working with multiples. And that’s not even including the core. Although only a small portion is heading towards informational technologies, we’re still talking about billions just in the Czech Republic.
At the same time, however, there’s only a limited number of people capable of working in IT. Meanwhile, there are plenty of projects waiting for us. So, it’ll be crucial to do them intelligently and to build on the current, existing, and standardised products, so we can limit the amount of work spent on them.
Alongside these projects, certain fields will surely open for completely new products. For example, at Unicorn, we established ourselves in the energy market during the first wave of liberalisation. At that time, the transmission system in the Czech Republic rather progressively separated from the rest of the sector and, at the same time, we developed the Damas trading system, which is the most-used system of its kind in Europe.
And will this involve primarily network balancing, production and consumption forecasting, and cybersecurity, or something else?
I really can’t say that a single topic is the main one. There are multiple problems and they’re all related. Instead, you could say which is the most urgent. Balancing, or rather ensuring the power balance, is definitely a big topic. And it will only grow in importance and size depending on how we end up retiring the rest of our coal fleet. It’s going to be quite difficult, but it’s safe to say that all key players are getting ready for it.
Smart metering is another thing whose installation runs in all large energy sectors. In the future, it will enable us to work with the system like never before. Of course, cybersecurity is also an important topic. Its significance continues to grow, but it isn’t something that can be solved by a single product that we’re currently building or implementing. It must be a continual issue; security aspects must be taken into consideration during all projects and activities.
However, we don’t want to go overboard because here in Europe, we sometimes tend to create rules that effectively prevent us from completing things. Whether that’s the GDPR, the AI Act, or others. We don’t want to be so secure that in the end, we can’t even make anything.
Will the IT world manage to catch up with the requirements for energy transformation, or is it actually already ahead?
It certainly isn’t the case that the world of information technology is holding back the development of the energy industry. If you consider how long it takes to build a power plant or even high-voltage lines, you’d be exaggerating to say that it’s an overnight thing here. At the same time, the IT world makes it possible to operate energy at the edge of what is even physically possible. But it won’t push beyond the physical limits. Thanks to IT, we can compensate for deficiencies in the infrastructure to some extent. But you can’t simply use IT to generate electricity.
In general, sophisticated management today is required at all levels. It starts with the trader, who’s now faced with a new situation, for example, considering that many consumers are also producers. And now sharing has been added to the mix. Of course, this reduces the accuracy of the predictions used. In addition, one can assume that the deviation price will increase and vice versa. In other words, the more you hit, the more you’ll earn. At the same time, it’s necessary to make more use of the regulation potential, especially at the low voltage level, meaning heat pumps, batteries, and similar controllable consumption. So far, it’s only at the pilot project level all around the world.
Today, we’re dealing with more things in Europe than in the Czech Republic, and we’re working to put ourselves in the position of the worldwide supplier of IT solutions for the energy industry. We actually just closed on a significant commission in Saudi Arabia, where the energy sector is also advancing.
Are they comparable when you consider they have their own oil, they probably have the AC running at all times, they have more even sun exposure…?
The two are comparable because the physical laws there are the same as here. Rather, the big difference is that their energy sector is much more centralised. Because they use more fossil fuels, their production is better regulated. But they too have plans for decarbonisation; the share of photovoltaics being made in Saudi Arabia should gradually reach as much as 30%. And this will require energy balancing as well. If they end up using batteries, hydrogen, ammonia, or something else, only time will tell.
Nevertheless, our project deals with managing the emerging regional electricity market in the entire Persian Gulf area, where many parallels can be seen with the historical stages of building the common European market.
Can such a fragmented and unstable system like the one in Europe even be regulated at all?
It can always be regulated. Worst case scenario, a part of consumption or production will split off.
But that’s a rather expensive solution.
Of course. But I’m also not saying that it has to happen. But an example like this must exist as an extreme scenario, and, if necessary, it must also be used. The Czech transmission system can absorb roughly 13 gigawatts of photovoltaics, and there’s no point in building more here. That’s because when the sun shines, there will be nowhere to store the electricity and the sources will simply turn off.
Won’t it end up so that households will also become your customers so that they can take advantage of price fluctuations?
Not ours, we focus more on the big players. We mostly specialise in everything related to the trade, whether it’s electricity or gas. But surely someone will make software for households. At the same time, our systems will enable the market, the transmission system, traders and aggregators, so that your heat pump can be involved in balancing, for example.
How does AI play a role in all of this?
That depends on what you consider artificial intelligence. In recent days, that term has gone hand in hand with ChatGPT. The use of these applications is relatively limited in our solutions. Except for user help, perhaps. If we’re talking about neural network technology as such, then that’s something different. We experiment quite intensively with these matters. AI is not self-saving, but its ability to self-learn is valuable. So far, however, it turns out that if you have an accurately programmed physics model, it works better. But if the circumstances change, you already have to readjust it, while the AI can adapt. However, to be able to do this, you need to have quality input data, which often presents both a problem and another opportunity for our services.
Energy trading advances quickly in real time, meaning you have to know exactly whether the network will enable the transaction to be completed, for example. This brings us close to the actual managing of the system, and there’s room for AI here as well, primarily from the perspective of optimising the space for completing trades.
But for now, it’s only at the level of providing suggestions. I can’t imagine letting AI manage critical infrastructure because it’s still just too unreliable.
And are we able to control such systems?
We have to.
But they’re faster than us.
And that’s exactly why their application options are limited. I’d be afraid to let AI physically manipulate network components. It must always be first checked by a human, a dispatcher. We’re actually working on a similar system right now for the Swiss; a system that can calculate thousands of possible scenarios for system configurations and locate the optimal solution. It’s going to be a unique tool, but there will always be a human factor here as well.
In general, this all reflects the fact that we started to replace a simple system with a complex one. Therefore, you will need to have a system that will manage it, another system that will control the managing one, and another one that will control the controlling one. But it will always be easier and faster to make a new IT system than to build a new line and a new power plant. However, of course, operating such a complex unit will not be cheaper than what we’ve had so far.
And the Chinese and Americans affirm that within 10 years, they’ll start supplying electricity via fusion. And that would be a stable source.
I won’t count on that so much just yet. Today, for example, there’s a lot of talk about small modular reactors, but even those really only work as prototypes. The shift from stable to intermittent sources has already occurred and will continue. We have to prepare for it.
