In Europe, the energy revolution is happening right now. 38% of energy is already sourced from renewable energy, surpassing fossil fuels for the first time. We expect this share to grow in the next few years, but for coming close to 100% renewable energy in the foreseeable future, some hurdles must be crossed.
One major limitation, which causes industries to stay skeptical is that renewables depend heavily on environmental conditions for energy production. Some power-intensive industries need a constant energy supply, which seems to be only provided by classical energy sources like coal, fossil fuels or nuclear power plants. In the FlexEuro project, an international alliance of industries, universities and governments are already about to solve this issue by prototyping flexibility concepts for power-intensive demands in real life.
The amount of power generated by renewable sources such as wind power or photovoltaics fluctuates due to changing environmental factors such as the weather and climate. It can’t be adjusted to the demand or the market price. Through new technologies, however, it is possible to create some flexibility on the demand side as well. This flexibility can be used to benefit from periods of energy surplus with favorable electricity prices while mitigating bottlenecks. Energy suppliers and big industries thus use load control demand-side management on the consumer side to make energy demand more flexible.
Increasing demand flexibility means giving the supply side more autonomy in deciding when to provide power to its consumer. While a flexibility framework guarantees to fulfill all power needs, the energy cost of all parties sinks due to the optimized energy distribution within the overall system. Flexibility on the demand side is an important success factor for the long-term operation of sustainable energy systems. As the optimization gets better with scale, power-intensive industries in particular have the potential to benefit from this economically. To reach the full potential, all relevant variables need to be modelled into a mathematical framework as in the FlexEuro project.
Decision Trees is involved in a research project sponsored by the German Federal Ministry for Economic Affairs and Energy. Joining forces with Fraunhofer ITWM, the university of Duisburg-Essen and TRIMET Aluminium SE, the FlexEuro project sets out to develop a decision support system for capitalizing on flexibility in power consumption. The basis for this research project is an aluminum production site from TRIMET, which was converted into a ‘virtual battery’. By implementing heat exchangers, this facility acts like a huge and intelligent electricity storage unit. 120 furnaces of an electrolysis hall are now prototyping the new framework for decision-making on an industrial scale. At around 1,120 megawatt hours, the ‘virtual battery’ alone has the storage capacity of a medium-sized pumped storage power plant - all with an efficiency of up to 95%.
The furnaces need a constant energy supply for the critical process of aluminum production. Despite being supplied by unsteady energy, TRIMET is able to keep the operating temperature stable. This flexibility of energy demand is achieved by developing and implementing controllable heat exchangers for the electrolysis furnaces. For up to 48 hours, TRIMET is now able to vary the energy supply from 75% to 125% without interrupting aluminium production.
This is all possible due to a completely new process control concept with innovative measurement and control technology as well as setting into place extensive process models to optimize the operating behavior of the electrolysis under flexibilization conditions. This innovative electricity storage facility makes it easier to integrate the unsteadily generated electricity from renewable sources into the power grid. At the same time the high rate of flexible power in this plant allows drawing cheaper energy from an intraday trading market and this way minimizing the overall energy costs.
Once the hardware is set into place, the focus can be shifted to the operational marketing of flexibility. The optimization tasks are constrained by the technical restrictions of the electrolysis process and electricity trading activities. Intraday trading is becoming popular as many players enter the market and the traded volumes continuously increase. At the same time the market is getting more complex, creating a rising demand for algo traders. Since these are built for taking into account numerous restrictions of complex models, it stands to reason to market the flexibility by using intelligent algorithms.
To realize an optimized trading strategy, the intraday trading market is represented by quantitative models and algorithms. The properties of the industrial plant are mathematically modeled as well. Feeding the mathematical models and algorithms with real data, they generate a set of actions. Based on the markets’ intraday trading prices, they calculate the best time frame to draw electricity and when to efficiently supply the system of consumers. Considering the technical feasibility of load changes in favor of greater freedom in energy trading results in overall cost savings. Besides revolutionizing aluminum production itself, this set of measures makes the case for introducing demand-flexibility into energy trading. This results in a significant contribution to the shift towards sustainable energy sources.
One of the key challenges for the FlexEuro project team is mapping the battery in all its complexity, but they can rely on Decision Trees’ many years of experience in optimizing different kinds of power systems. Decision Trees take part in developing quantitative models and algorithms by sharing their expertise in modeling such highly complex systems. A focus lies on the investigation of the added value of stochastic optimization in the area of short-term electricity trading.
As we just explored, increasing flexibility in energy consumption is a huge potential for energy intensive industry plants. As renewable energies themselves provide unstable energy, the demand-side should catch up. The investment to make the consumption more flexible opens doors to intraday-market trading using stochastic optimization by Decision Trees and is rewarded by overall cheaper energy costs. A better bottom line paves the way for more developments in regards to sustainable energy and can increase their overall share in the long run.