The Virtual Power Plant: Enabling More Electric Power with Fewer Emissions

Lawrence Willey

Digital models of power plants are showing the way to lower emissions. Other benefits include traditional measures of availability, reliability, and performance. Big-data analytics provides more, including flexibility, maintainability, and sustainability.

Everyone is becoming more aware of the groundswell of activity and improved results from the Industrial Internet of Things (IIoT). The virtual power plant is one advancement that's beginning to make significant strides in terms of lowering emissions and improving profitability for fossil-fired power plants.

The Virtual Power Plant Is a Digital Model

Power plants produce electricity that enables our world as we know it today. In the quest for ever-improving performance, what better place to focus—using big data and advanced diagnostic analytics—than the very machines that make this possible? Today's digital power plant is a virtual model (i.e., Twin) that incorporates sensors, more data (with ever-smaller time steps), fleet perspective learning, and precise integration of operational and external data sets. It quickly produces actionable information that supports engineers designing new or improved products, as well as an asset manager's ability to make better decisions for today's operating equipment.

Like all large machines, heavy-duty gas turbines have design operating conditions that improve performance and lower consumption or generation of secondary elements. For natural gas-fired plants, consumption or generation would be for the natural gas fuel and the combustion exhaust products, respectively. Finding ways to lower emissions as operational demands change is key for ensuring that the primary goal of electricity generation and delivery is achieved in the most efficient way possible. The big-data digital model is at the heart of the virtual power plant and makes this possible.

Connected Power Plant Models Share Greener Operating Profiles

Increasingly, natural-gas-fired combined cycle power plants are being dispatched more often and on shorter notice compared to historical operation. For each power plant, data collected every day during actual operations informs its digital model. This results in rapid insights for improved capacity, efficiency, and emissions for the plant. The information captured by the digital model for power production and emissions yields key insights much sooner than was possible before.

As more and more of these power plant models are created, machine learning within each of the unique circumstances surrounding these individual plants begins to yield additional insights and improvements. Digitally connecting these models and applying more loops of machine learning will, in turn, help accelerate the best results for the overall group. This then leads to better new plants, as they are added to the overall system. The end result is better intelligence sooner, which informs the asset manager regarding how to limit emissions and increase profitability, while yielding actionable information that will avoid costly downtime.

How Does It Sound to Have Quicker Decisions and a Better Understanding of Potential Outcomes?

As the virtual power plant can run nearly endless digital scenarios of almost every possible outcome, the asset manager now has the ability to weigh the outcomes from a range of decisions about the power plant, other generators in the system, and the overall fleet. With this information produced in summary format and quickly and continually refreshed as input parameters or assumptions change, the asset manager has the confidence to make the right decisions and remain in control.

For design engineers, picture having precisely time-stamped and streaming data for the actual working oil in the gearbox of an operational wind turbine. Besides the basic pressure and temperature information, this new technology includes viscosity, density, dielectric constant, water contamination, and very small particle counting, among other metrics. That last measurement doesn't just include larger particle (chip) detection like that employed in other condition monitoring systems (CMS). Rather, this new IIoT product detects and presents the actual binned counts of <4-21 micron particles for each and every 30-second data snapshot in real time. Imagine the incredible insights that can quickly be obtained for how wind conditions and wind turbine operational parameters influence minuscule particle generation inside the gearbox.

With this information, the design engineer can ensure the very best form and material for improving new or rebuilt gearboxes. An asset manager now has the information he or she needs to obtain the most energy production, while operationally throttling the individual turbines, as needed, to meet lower-cost repair or replacement windows.

Early Adopters of This Technology Will Be Rewarded with Better Profitability

The IIoT is rapidly being developed across a broad spectrum of equipment and infrastructure, especially power plants. Big data and predictive analytics—for subsystems and across the larger power plant systems—are yielding significantly improved power plants. The financial benefits of investing in digital twin technology are lowering emissions with more energy production and less wear and tear on the equipment.

As a result, the big-picture enterprise system of systems—traditional systems engineering combined with the enterprise activities of strategic planning and financial analysis—and the delivery of electricity to power society are rapidly improving. Operators and equipment are being better utilized, and improved profitability is enabling the twin power trend to self-reinforce and accelerate. Those who quickly recognize what is happening in this arena will reap early and increasing rewards.


Wind energy continues to become more competitive, expanding its share of global energy production.

The growth and intersection of four energy trends—decarbonization, decentralization, vehicle electrification, and energy access—is transforming the power sector and society.

Will the responsibility of intermittent energy source balancing fall primarily on batteries and energy storage, or can digitized thermal power plants take on some of the task?