Auto Optimize for Tower Field Layout gives wrong results

Paul:

When I run a tower case with a maximum receiver flux of 2000 suns, I get a substantially different answer (LCOE) depending on whether I manually press the field layout and then the optimization buttons, or check the boxes for automatic (layout and optimize every time). In watching the optimization, it is clear that when the Always Layout and Always Optimize boxes are checked, the optimization goes to a 1000 suns limit on the Receiver, rather than optimizing to the requested 2000 suns.

If I manually click the buttons for layout then optimize, it clearly goes to 2000 suns.

Further, if I set it for 1000 suns, and then manually optimize, the answer is slightly different than the auto-optimized 2000 sun version, which means the auto-version is not only reverting to 1000 suns, but appears to also be over-riding another setting of mine that I have not yet determined what is over-written.

These issues are also seen during parametric studies in which I set the max flux to 2000, and turn on optimize and layout always, and then vary something like slope error. Clearly optimizing to 1000 suns. This pretty much precludes doing parametric studies.

I am attaching an example file. Please try single runs with both manually (buttons) and automatically (check boxes) Layout and Optimize.

Thanks
Chuck

Dear Chuck,

Thanks for this. We are investigating and I will follow up when we know more.

Best regards,
Paul.

Thanks, I appreciate you getting back to me. Just got off the phone with our customer for this project. I am thinking that maybe I need to learn the scripting language to bypass these optimization issues? Would I be able to "manually" press the optimization buttons from a script, and thus build my own "optimize every time" capability? Otherwise, I am stuck with single runs and manually optimizing, which severely restricts the parameter space that can be explored. Might be worth me trying due to tight schedules on this project.