German scientists bundle renewable energy sources to form a ‘virtual’ power plant
Fine-tuning a difficult balancing act
The institute’s plan — to generate reliable electricity output from distributed generation, the majority of it intermittent — requires precise monitoring at all points in the system. Turbines can be rotated into or out of the wind to deliver more or less power. Solar photovoltaics deliver their power during the day, when the wind is most subdued, meaning the system’s two intermittent power sources are largely complementary.
And for those hours when neither sun nor wind is yielding a bounty of electrons, biogas can be brought online to keep output constant. When the intermittent renewables kick back into gear, that gas can be stored for future use or converted into heat.
At the center of this energy network, a software platform monitors each source of energy and adjusts accordingly, ensuring that the system maintains a steady output.
Because these resources function in concert to produce power equivalent to conventional baseload, the researchers term their conglomeration a “virtual” power plant.
Nor is the work purely theoretical. The institute has already tested its software with real infrastructure, balancing the loads of distributed sources to produce a steady output of 80 megawatts.
“The results of [our] Combined Power Plant 2 project demonstrate that network reliability can be guaranteed even when relying purely on renewables,” Rohrig said.
Similar ideas involving distributed generation and a centralized energy management system have been explored in the United States. Sometimes called “virtual utilities,” these systems have employed technologies like combined-cycle gas power, various types of energy storage and more traditional renewables.
In addition to achieving high penetrations of clean energy, these systems could potentially be more flexible and better able to cope with unexpected challenges, such as those posed by extreme weather, than conventional baseload power plants.
Tailoring demand to meet supply
On the supply side of the equation, researchers at the Franhaufer Institute’s Factory Operation and Automation branch are working out a better balance for the power that customers consume.
“We need to change our thinking from the now-common generation of power geared toward consumption to consumption geared toward providers,” said Przemyslaw Komarnicki, part of the operation and automation team.
Whereas the team behind the virtual power plant is striving to provide on-demand power, Komarnicki and colleagues are looking for ways to meet them halfway, designing software to automatically adjust power consumption during periods of high or low electricity supply.
Their energy management system offers a real-time relay between a building’s control system and attached renewable energy systems and storage — a solar panel on the roof, for instance, or a battery array in the basement — as well as a fleet of electric cars parked and plugged in the building’s garage. The system supplies power to the building’s heating, ventilation and air conditioning needs first, then uses whatever excess is available to charge the vehicles.
Factoring in meteorology and expected work traffic, neural networks make regular, quarter-hourly forecasts, which the system uses to optimize its energy use.
As with the virtual power plant, the system relies on a certain level of surplus power that can be generated, stored and dispatched as needed. The cars themselves could be a site for energy storage, functioning as large batteries during the day while their owners are at work.