Concentrating solar power tower at the National Solar Thermal Test Facility (NSTTF)

Concentrating Solar Seminar Series to examine advances in metrology, deep learning-based airborne monitoring methods

August 22, 2023 2:03 pm Published by

Join Concentrating Solar Power (CSP) manager Margaret Gordon and CSP researcher Randy Brost on Tuesday, Aug. 29, 2023, from 10:00 a.m. to 11:00 a.m. MT as they host the next installment of the Sandia National Solar Thermal Test Facility Concentrating Solar Seminar Series. This series’ installment will focus on advances in metrology and standards for central receiver systems, and deep learning-based airborne monitoring methods for heliostats in solar tower power plants.

Speakers scheduled to present are:

Marc Röger: Advances in Metrology and Standards for Central Receiver Systems — A German View

Concentrating solar technologies (CST) use mirrors to concentrate the sunlight to generate high temperatures for electricity generation, industrial heat production, or solar fuels. Metrology in CST helps to identify problems and to iteratively optimize individual components (mirrors), component assemblies (heliostats) or whole systems (heliostat field or solar plant). In this talk, focusing on recent advances, Röger will give an overview of metrology techniques and respective standard developments that have been identified from the German Aerospace Center (DLR) as essential for the sustainable development of central receiver systems.

Marc Rögers

Marc Röger, Ph.D. in mechanical engineering from the University of Stuttgart, is a senior researcher leading the ‘Qualification-Systems’ team at DLR. He has 24 years of research experience in CSP. Since 2005, he has been a part of the permanent DLR delegation at the Plataforma Solar de Almería in Spain, working on qualification of collectors, heliostats, and receivers. He leads the SolarPACES “heliostats” working group (Task III) and is active in the International Electrotechnical Commission (IEC)-TC-117 62862-4-3 working on design qualification of heliostats. He has been associate and guest editor of the ASME Journal of Solar Energy Engineering and is the author of several scientific publications and patents. Heis a co-founder of the CSP Services company. Marc and his team have received several technology and scientific awards.

Rafal Broda: Towards Deep Learning-Based Airborne Monitoring Methods for Heliostats in Solar Tower Power Plants

While deep learning methods have proven their superiority over conventional algorithms in many domains, their use in image-based heliostat monitoring remains limited. Rafal and his team aim is to bridge this gap by developing models to solve various detection tasks in this field. Their approach is distinguished by the creation of synthetic image data for training, which grants them access to vast amounts of data essential for developing effective deep learning models. Moreover, this approach enables them to simulate operationally relevant scenarios across a diverse range of situations difficult to replicate using real data. Broda will present their initial findings and current progress.

Rafal Broda

Rafal Broda holds a double master’s degree in mechanical engineering with specialization in energy engineering from RWTH Aachen University and Tsinghua University in Beijing, China. Throughout his academic journey, Broda focused on the field of machine learning and worked at various research institutions to deepen his knowledge and expertise in this area. During his master’s thesis at the German Aerospace Center (DLR), he developed a deep learning-based correction for cloud camera-based solar radiation measurement.

Currently, Broda is pursuing his Ph.D. at the Institute of Solar Research at the German Aerospace Center (DLR), where he is applying his knowledge to develop AI-based solutions for the airborne qualification of solar thermal power plants.

To learn more about the National Solar Thermal Test Facility Concentrating Solar Seminar Series, contact Margaret Gordon.

Sandians are welcome to attend in-person at 9981/146 or attendees may join via Teams.