Session: 05-13 Heliostat Consortium 3

Paper Number: 114008

114008 - An Indoor Reflected Target Optical Measurement System for Concentrated Solar Power Mirrors 

Accurate optical surfaces are a primary driver of concentrated solar power plant performance. Errors in mirror pointing and tracking, the canting of individual mirror facets, and the surface slope of the mirror itself can be caused by errors during assembly, transportation, wind loading, gravity, and many other sources. The tools that exist to measure these error sources today largely rely on fringe deflectometry (SOFAST, QDec, others), or photogrammetry with targets attached to the mirror surface.

In 2022, NREL presented a new indoor optical measurement system to measure fully assembled heliostats during development and on the assembly line for quality assurance purposes, which has undergone significant development in the last year, including the installation of a new, much larger laboratory setup.  This system, called ReTNA (Reflected Target Non-intrusive Assessment), uses a printed pattern as a deflectometry target. Using printed pattern features as deflectometry targets instead of fringes allows the software to measure the surface slope in two dimensions with a single target. This eliminates the need to for targets to change, as with fringe pattern targets, and therefore the target for the target to be a large screen or projected image. In the last year, NREL’s ReTNA tool has been adapted from the large, wide targets used in fringe deflectometry (a ~5x5 meter projector screen) to use a tall, thin target (less than 1m wide).  In this new setup, the target only spans one dimension of the mirror. A camera moving on a simple, automatic track system pans the reflected target across the mirror surface to generate a full optical surface model of the mirror. This new setup drastically improves the usefulness of the ReTNA system. For example, a single stack of lightweight modular printed targets affixed to a wall (<1m wide by 8m tall) should be able to effectively measure a 4x4 meter heliostat in ambient laboratory lighting. With a traditional fringe system, this would a 8mx8m projector screen, and carefully controlled lighting.

We will present these new developments, including the new setup, target patterns, software methodology for fast, automatic optical analysis, and important lessons learned throughout the process. ReTNA results will be compared to established tools to establish the uncertainty of this new method.

ReTNA leverages recent improvements in computer vision and photogrammetry to remove as much careful calibration from the setup process as possible, making the tool low cost, simple, and portable.  While it will not achieve the resolution of a fringe deflectometry system, its advantages; drastically lower cost and simplicity of setup and calibration, make it well suited for CSP evaluation, and allow it to fill a different industry need. We intend for this tool to be used as a low-cost method for heliostat developers to test new, fully assembled heliostat designs at a variety of orientation angles and simulated loads, lowering the barrier to entry for innovative heliostat development. We also hope to one day integrate this into heliostat manufacture, for assembly line quality assurance and control.

Presenting Author: Devon Kesseli National Renewable Energy Laboratory

Presenting Author Biography: Devon Kesseli is a researcher at the National Renewable Energy Laboratory, working on optical, thermal/fluid, and economic modeling for concentrated solar and geothermal technologies. He received his MS in Mechanical Engineering from University of Colorado, Boulder, and a bachelors of physics from Reed College.