GE Research: flexible plants for more renewable intensive grids

With 40% of the world’s electricity and 30% of electricity in the US produced from coal, it remains a large and important part of the energy landscape. This large installed base delivers reliable and stable electricity which can be improved in ways that reduce local and global emissions, including greenhouse gases. With the rapid growth of renewable resources and changing realities of the power market, coal can play an important role as a stabilising force on the grid.

“These projects represent an incredible opportunity for our teams to advance technologies that can help coal plants across the US and beyond operate with cleaner, higher performance and support the ongoing growth of renewable power,” said Dave Begley, General Manager for GE Steam Power in the Americas. “We’re appreciative to the DoE for providing the funding to move these projects forward, and excited about the possibilities that can come from blending the power of hardware and software innovations in coal plant operations.”

The two projects involving the development of AI and digital technologies will be led by teams of engineers at GE Research in Niskayuna, NY. They are:

AI learning system for coal-fired power plant fault detection and diagnosis

Some scientists and engineers at the company will use historical data to model a normal operating plant and train the system to recognise and provide an early detection and root cause analysis to operators if the physical plant operations fall outside of the normal operating parameters. The idea is for faults to be detected and fixed early to mitigate any potential disruptions and improve the overall reliability of plant operations.

“We will be employing AI, advanced controls and other digital capabilities to enable coal plants to more quickly change its load profile to balance fluctuations with intermittent renewable energy generation,” said Mustafa Dokucu, a Senior Scientist leading the model-based estimation and predictive controls project using digital twins. “By transforming coal into a more flexible power source, it will help firm up grids with a more renewable intensive energy mix.”

DoE Funding: US$1 999 853

Model-based estimation and predictive controls with digital twins to improve flexibility, reliability and efficiency of coal plants

Drawing from an extensive model library of coal plant components, GE engineers will create digital twins, or digital replicas of these components to estimate and then optimise the heat rate of plant operations for improved efficiency. Each point of efficiency gains translates into 2% of CO₂ emission reduction. For context, The company’s digital twins are continuously learning models that update as new data is collected and analysed from sensors on the asset or components themselves, fleet data of other similar components, human experts, and from simulations that examine ‘what if’ scenarios.

In addition to improving efficiency, GE digital twins of coal plant components will be utilised to optimise efficiency and deliver 30% faster ramp rates to more capably handle the intermittency of renewables. Finally, the Twins will be used to determine the health of these components for improved reliability. Feng Xue, a Senior Scientist leading the AI learning system project added, “As we build in more flexibility, we also will use AI to improve the reliability of coal plants themselves. The AI learning system we’re creating will be trained to detect faults and identify their causes before they lead to any issues.”

DoE Funding: $1 999 060

The project involving new system development for the boiler will be led by the GE Steam Power team in Windsor:

Plasma ignition and combustion stabilisation technology to improve flexible operation, reliability and economics of an existing coal-fired boiler

A third project will be led by GE Steam Power business in Windsor, Connecticut, and focus on the development of advanced plasma technology to improve coal power plant flexibility by enabling low load operation, cold starts and faster ramp up time while lowering operating costs. The goal of this project is to fully integrated and demonstrate the technology in a field proven system, so it could then be made commercially available for other coal-fired power plants.

DOE Funding: US$3 615 340

Read the article online at: https://www.worldcoal.com/exploration-and-development/01072019/ge-research-flexible-plants-for-more-renewable-intensive-grids/