The element of insecurity has increased, simultaneously amplifying the importance of forecasts. Whoever handles them the best, earns the most. At Unicorn Systems, we help companies predict the production of Renewable Energy Resources (RES) using our Lancelot FMS solution.
Last year, RES accounted for 44.7% of electrical energy production in the EU, awarding them the top spot for the first time in the European electricity framework. Not only does this constitute a significant result of the effort to reduce emissions but it’s also an important sign of how the energy market is changing. The growing percentage of solar and wind power plants is altering the rules of the game. Among many factors, the need is increasing to forecast this future development as best as possible, and the focus of these forecasts is becoming more and more diverse. And not only for the frequently declining network balance but especially for its economic effectiveness.
The energy industry we’re currently abandoning faster and faster was relatively simple in that multiple large, conventional sources supplied each given range of distributors. The energy flowed in one direction, from production to the consumer, whereas the first was manageable and the second by and large foreseeable. However, renewable sources began gradually mounting into frameworks, especially solar parks and wind power plants. These sources were larger at first before smaller ones also started popping up – all the way to solar panels on the roofs of family homes. Thus began, and still continues today, the decentralisation of production, while energy flow has become two-way. Also, the photovoltaics’ and windmills’ performance is dependent on the weather, another new element of insecurity. They can no longer be managed systematically; we have to forecast their production based on meteorological data, such as light exposure, wind, cloud cover, temperature, etc. And these data fluctuate around a certain level of probability as well. It should be noted, however, that RES aren’t some kind of loose cannon in the system as is sometimes claimed. Their production too can be regulated, for instance, by tilting the windmill blades or shutting down individual sections of solar power stations.
Of course, the introduction of RES and the other phenomena mentioned above comes with a role change for the energy market participants. Balance in the network is primarily a task for its operator and achieving it is more difficult than before. That’s why the relevant company needs to forecast as best as possible how the system will behave, when and where imbalances may occur, and by how much they’ll deviate. This provides an answer to the key questions of whether it’ll be necessary to proceed with balancing, what sort of flexibility services should be deployed, and how much it’s going to cost. Production predictions of increasingly prominent renewable sources have become a crucial quantity within this puzzle.
In another situation, the trader prioritises minimising costs and maximising turnovers. They also need to keep the purchased production within the contractual deviations from the actual consumption because, otherwise, they’ll have to pay a penalty, but it’s primarily in their best interest to purchase low and sell high. They can store the electricity and speculate using various methods. Now’s the time to mention another fundamental change to the energy market: electricity prices are in much greater motion than before. And again, that’s why the trader needs to forecast what will play out within his portfolio as precisely as possible.
RES compound this task. It used to be the case that if it supplied a given range of distributors, then the factory had a standard set of consumption for the period, the shopping centre consumed some, and the neighbourhood of family houses consumed some too. All one had to do was add everything together, buy the total volume, deliver, and then collect. Today, the factory and a portion of production houses have photovoltaic panels on their roofs, and these consumers are more or less self-sufficient, depending on the current weather. Plus, they can store energy in batteries and remain self-sufficient even in times of less-than-ideal weather. Therefore, they no longer need energy from the network all day long but only at certain times. For a trader to work economically, they need to know what sort of demand to expect with at least some level of clear likelihood.
In short, forecasting is the key to effective operations in the energy market. And that’s true even though they can’t be set with such relative simplicity as in the days of the classic, centralised energy industry. They must be more dynamic, more flexible as an overall industry to respond to the needs of various players, while also being able to absorb a notably larger amount of data and variables. Also, they can’t rely on a so-called black box to make all their predictions. Companies today need to be able to play with forecasts and model out various scenarios.
As the industry itself grows more complicated, so too does the role of those developing software and predictive energy models. Space has opened up for today’s highly worshipped AI, which, at least for now, isn’t capable of forecasting with greater quality than the current statistics and mathematics models but can better handle some other kinds of tasks. For instance, retraining models according to developing realities, identifying trends, segmenting client portfolios, and time-consuming data preparations and new variable implementations. Using AI helps create forecasting models faster and more effectively, which benefits not only the creators but especially their clients.
However, at Unicorn Systems, the partner of energy companies all across Europe, we rely first on our many years of experience, the expertise of our teams, and the continuity represented by our energy market solution Lancelot. We develop it in response to our clients’ ever-changing needs, and forecasting is at the centre of our interests. The ability to better predict the final outcome always results in money either earned or saved. Of course, this depends on the size of the specific company, but we’re not talking in the thousands but rather in the millions of earnings and savings every year. For some of the biggest players, the numbers easily reach the hundreds of millions.