Session: 18-01 HelioCon Metrology

Paper Number: 142437

142437 - Moving Nio to Commercial Ready 

Abstract: 

A common issue that occurs on heliostat fields is the misalignment of heliostats from their central power tower. This causes a loss in energy production and money for these fields. Misalignment of heliostats can be classified as optomechanical errors which can arise from various issues such as mirror surface deformations, manufacturing and installation errors, mechanical defects, weather disturbances such as wind loads, or from cleaning and calibration procedures. However, searching for which of the heliostats on the field are misaligned is time intensive and expensive and once found physically testing their inaccuracy can result in greater misalignment. To combat this Non-Intrusive Metrology (NIO) was developed to characterize the optical errors of heliostats by using drones to capture images of the heliostats to survey the heliostat fields without perturbing the heliostats and allowing for the errors of the field to be measured faster and less expensively than can be done by people. This way heliostat field operators can track the efficiency of their fields while also being able to make informed adjustments to their heliostats to optimize the performance of their field. As NIO has left its initial testing stages and surpassed laboratory benchmarks its prototyped form is not suitable for commercial use, as it has several severe issues such as over utilization of memory, long operation times for even small data sets, and has a poor user interface. Before commercial use is possible NIO must undergo a transformation to make it a suitable product for widespread use that can run efficiently, effectively, and reliably. To achieve this NIO went through a series of upgrades. Throughout this process we renovated how NIO manages its data using Object Oriented Programing to reduce unnecessary memory allocation and to promote better memory utilization and quicker memory access, we needed to conduct sensitivity and stability analysis on the calculations for optical errors, and heliostat corner detection to ensure proper reliability, precision and accuracy for consumer use, cut down its operation times by implementing algorithms and methods that were less computationally expensive, and finally, we needed to implement a user interface that is easy to use. Through this we were able to make NIO more rigid, and easier to use as it transitions from its prototype laboratory form to broad scale commercial use. Allowing NIO to effectively handle larger data sets reducing its overall process run time by 33% and cutting down its memory utilization by 20%.      

Presenting Author: Kyle Sperber NREL

Presenting Author Biography: Kyle is a graduate student at the Colorado School of Mines studying Applied Mathematics and Wave Phenomena, and Physics. Kyle’s academic research includes studying nuclear radiation shielding for Martian habitats and designing an electromagnetic pulse shaper for ultra-fast imaging. In his free time, he enjoys hiking, trail running, and climbing.

Authors:

Rebecca Mitchell National Renewable Energy Labs
Kyle Sperber NREL
Tucker Farrell National Renewable Energy Labs