The energy demands of industrial companies can be gigantic and are often on a par with the consumption of small towns. Take MAN Energy Solutions at its headquarters in Augsburg, Germany where the entire facility consumes 55 to 65 GWh of electrical power annually, and loads regularly peak at 15 MW – equivalent to the daily consumption of 3,000 households. For public grids, such sudden peaks in load can be challenging to handle, which is why the company operates a 5.4 MW gas-engine power plant to shave such spikes off the grid.

“Other, large, industrial companies have the same challenges as we do, or face even more complex energy-management needs. Traditionally, such companies ran thermal-power plants but, given the need to decarbonize, the question arises as to how to integrate renewable resources into the mix and how then to eventually replace fossil-generation altogether,” says Michael Raila from MAN Energy Solutions, Executive Assistant to Head of SBU Power Plants.

Integrating rising shares of renewables into a system, however, poses its own challenges: “Industrial plants need a stable energy supply at all times but wind and sun can be intermittent, which makes energy management a challenging task. Another effect of such volatility is highly dynamic energy-pricing – at least in the EU,” adds Raila. “Accordingly, companies will have to find a way to secure their energy supply and make the right economic choices. They’ll also have to take decisions both in the near-term and repeatedly – over and over – 24 hours a day.” 

The ILLUMINE demonstrator: At MAN’s Augsburg site the project partners have installed a small-scale, hybrid power-plant to the energy-management system.

The team behind ILLUMINE (from left): Christian Wiethe (Fraunhofer); Mohamed Elsawaah (MAN Energy Solutions); Dominik Spaun (XITASO); Michael Raila (MAN Energy Solutions)

To come up with a solution to intelligently master future industrial energy management, MAN Energy Solutions initiated the ILLUMINE research project. It did so with a number of partners, namely software development company, XITASO and the Fraunhofer Project Group Business & Information Systems Engineering, which is associated with the University of Augsburg, the University of Bayreuth as well as the Augsburg University of Applied Sciences. “We are working on an intelligent control system that helps to operate complex power-systems and chooses the most-efficient energy source at any time,” says Raila.

Since the project’s kick-off in early 2018, promising progress has been made. Already today, ILLUMINE has become even more than just code and smart algorithms. At MAN’s Augsburg site, the project partners have installed a small-scale, hybrid power-plant consisting of an MAN gas engine, a photovoltaic system and a battery-storage unit to put things to the test. The test facility can simulate various loads and operate in grid-parallel or island mode, enabling researchers to simulate more than 20 different application scenarios. In this way, the project team can test how to optimally operate the system components, how quickly the engine can react to fluctuations from the photovoltaic system, and at what point it makes sense to buffer any shortfalls with battery power.

The core of ILLUMINE is an elaborate energy-management software.

At the heart of ILLUMINE is an elaborate, energy-management software, which is implemented by XITASO based on mathematical models from Fraunhofer researchers. “The ILLUMINE platform uses specifically developed, complex optimization-algorithms to control the energy supply and demand,” says Prof. Björn Häckel, who heads the project together with Prof. Hans Ulrich Buhl at Fraunhofer. “It’s able to process data from generators, solar-power systems, wind turbines, energy-storage devices, electricity markets and consumers for this purpose.”

For instance, in periods when electricity prices are high and weather conditions poor, ILLUMINE can fall back on power generation with gas engines or operate on battery-storage instead of drawing electricity from the public grid. Depending on the market price, it may even be economically viable to generate surplus energy and feed it into the grid.

Prof. Häckel adds: “We need a software platform that can not only forecast how the energy price and weather will develop in the short term, but also how production processes and the energy supply can adapt to it.”