Software experts weigh in on a new autonomously operating control system that uses real-time algorithms and forecasting tools to intelligently coordinate microgrids, hybrid power plants and energy parks – with top ecological results.
The demand for microgrids is booming. The latest market forecasts from Navigant Research predict that by 2028, the installed capacity will increase fivefold, from the current 5,000 MW to 25,000 MW. Pressure to decarbonize the power supply makes microgrids – along with other hybrid power suppliers like hydrogen power plants and energy parks – all the more attractive. Power plant and grid operators are no longer seeing this as simply an opportunity to stabilize their power supplies locally or to supply power to remote areas. “With an optimal combination of renewable energies, hydrogen electrolyzers, energy storage systems, waste-heat recovery, and hydrogen-capable power plants, we’ll also be able to substantially reduce the proportion of fossil fuels in the power and heat supply,” explains Dino Ablakovic, who is in charge of the microgrid product portfolio at Siemens Energy. He’s an electrical and software engineer familiar with the requirements and challenges of hybrid power supplies based on his experience with over 50 microgrid projects in which he served as solution architect. “But it requires a smart controller,” Ablakovic adds. “It only works if the controller can autonomously control the fluctuating availability of renewable energies and fluctuating demands of consumers in real time.”
The Omnivise Hybrid Control digital control system from Siemens Energy, which Ablakovic worked on with in-house experts and external expert partners to advance its development, can do exactly that. As a modular, standardized platform with flexibly scalable applications and tools, it’s capable of autonomously coordinating various combinations of power supply components around the clock and optimizing the efficiency of their interactions both economically and ecologically. The controller ensures that output is balanced and that frequency and voltage are controlled in the hybrid system – and not just in standard operation, but also in exceptional situations like load shedding, peak shaving, black starts, and resynchronizations following a power failure in the grid. Comprehensive development work performed over the past few years laid the groundwork and turned the Siemens Energy SPPA-T3000 control system into a smart, adaptive microgrid brain with real time-capable algorithms and forecasting tools.
Two experts in particular from the Siemens Energy development team made a major contribution to the innovations that are incorporated in Omnivise Hybrid Control: Daniel Stierhof, an expert in real-time algorithms and Achim Degenhardt, an expert in the simulation of complex industrial plants. They’re the ones who ensured that the necessary real time-capable algorithms can be developed in a simulation environment that realistically simulates complete microgrids and hybrid power plants. “Real-time capability places extremely high demands on algorithms,” explains Stierhof. “It can be achieved only if each intermediate result in the development phase can be directly tested in simulations to determine how the relevant control commands operate in reality. The larger question is always whether the components that are being addressed respond at the right time.” “A lot depends on how real the simulation environment is,” adds Degenhardt. “We need excellent simulation experts who can generate digital twins of complex energy supply systems, either in their own team or in our network of research and development partners.”
With Omnivise Hybrid Control, hybrid power plants have more than just an economically and ecologically smart control system. The integrated Dispatch Optimizer also enables interesting additional business with a plant’s own hydrogen production, storage, and reconversion. These power plants can take advantage of periods in which there is an oversupply of cheap or even negatively priced solar and wind power to produce and store hydrogen. They can then reconvert the hydrogen to power and feed into the grid at a profit during periods of high demand and prices – typically in the evening hours.
The algorithms they developed not only establish the conditions for autonomously controlling a complex mix of different energy supply components in standard operation or during load shedding and resynchronization. They also make it possible to use an energy management system that optimizes the components’ operating sequence based on boundary conditions like weather, profitability, carbon footprint, and availability. Every 15 minutes, the Omnivise Hybrid Control system’s Dispatch Optimizer calculates a new, optimized generation schedule for the next 24 hours based on weather forecasts, energy market prices, demand values, and other information. The schedule is then transferred to the controller for the hybrid power supply and is processed there. “Naturally, it’s always done in such a way that the grid remains stable no matter what,” says Ablakovic. “For example, if an unexpected cloud suddenly reduces solar power or if electricity consumption unexpectedly exceeds the value in the schedule, the controller immediately and autonomously counteracts this by feeding in maintained operating reserves and starting a generator to supply additional power.”
On the Galapagos island of Isabela, a hybrid power supply system comprising a photovoltaic plant, biofuel generators, and batteries has been controlled by a prototype of today’s Omnivise Hybrid Control system since 2018. As a result, the amount of fossil fuels currently required for power and heat generation on the island has been reduced by 500 tons, and CO2 emissions have been reduced by almost 2,000 tons annually. More deployments of the control system are currently being readied for an Australian mining company’s microgrid and for a municipal energy park in Germany. Both customers are aiming to fully decarbonize their energy requirements over the next few years. The development team at Siemens Energy is currently working on more innovations for these use cases. “Among other things, that involves real time-capable algorithms for edge devices that allow us to directly control the loads of individual power consumers – for example, heat pumps and cooling systems – using Omnivise Hybrid Control,” says Ablakovic. “It’s an extremely important step forward in the standardization of microgrids and energy parks, especially in terms of the communication between combinations of components from different manufacturers. It’ll permit more plug-and-play, will speed up the engineering and implementation of these types of solution, and will reduce project costs, all of which will make its deployment even more attractive to potential operators and their financial backers.”
“Self-learning algorithms, meaning AI, will also play an important role in Omnivise Hybrid Control in the future,” says Stierhof. “Their potential for improving the reliability of the weather forecasts we use for the energy management system is tremendous.” Simulation experts from the network of research and development partners will also contribute to achieving this development goal. In this particular case, this includes the RWTH Aachen University’s Institute for Power Electronics and Electrical Drives.